<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>SourceScore VERITAS — verified AI/ML claims</title>
    <link>https://sourcescore.org/claims/</link>
    <atom:link href="https://sourcescore.org/claims/feed.xml" rel="self" type="application/rss+xml"/>
    <description>Signed, sourced, citable AI/ML claims. Each item has 2+ primary sources and an HMAC-SHA256 signature. New claims are added regularly; subscribe to follow.</description>
    <language>en-US</language>
    <lastBuildDate>Sat, 16 May 2026 15:16:06 GMT</lastBuildDate>
    <generator>scripts/generate-claims-rss.mjs</generator>
    <item>
      <title>Transformer architecture introduced in paper: Attention Is All You Need (Vaswani et al., 2017).</title>
      <link>https://sourcescore.org/claims/ad17e76a8baad7a1/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/ad17e76a8baad7a1/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Transformer architecture introduced in paper: Attention Is All You Need (Vaswani et al., 2017). Verified 2026-05-16. Confidence 100%. 3 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/ad17e76a8baad7a1.json.</description>
      <category>transformer, attention, foundational, vaswani, 2017, nips</category>
    </item>
    <item>
      <title>Reinforcement Learning from Human Feedback (RLHF) introduced in paper: Deep Reinforcement Learning from Human Preferences (Christiano et al., 2017).</title>
      <link>https://sourcescore.org/claims/67866330cd60e54d/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/67866330cd60e54d/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Reinforcement Learning from Human Feedback (RLHF) introduced in paper: Deep Reinforcement Learning from Human Preferences (Christiano et al., 2017). Verified 2026-05-16. Confidence 100%. 3 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/67866330cd60e54d.json.</description>
      <category>rlhf, alignment, foundational, christiano, 2017, nips</category>
    </item>
    <item>
      <title>Retrieval-Augmented Generation (RAG) introduced in paper: Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks (Lewis et al., 2020).</title>
      <link>https://sourcescore.org/claims/d15057ced937a103/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/d15057ced937a103/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Retrieval-Augmented Generation (RAG) introduced in paper: Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks (Lewis et al., 2020). Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/d15057ced937a103.json.</description>
      <category>rag, retrieval, foundational, lewis, 2020, nips, facebook</category>
    </item>
    <item>
      <title>Low-Rank Adaptation (LoRA) introduced in paper: LoRA: Low-Rank Adaptation of Large Language Models (Hu et al., 2021).</title>
      <link>https://sourcescore.org/claims/d7b97d1b93d8d8bc/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/d7b97d1b93d8d8bc/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Low-Rank Adaptation (LoRA) introduced in paper: LoRA: Low-Rank Adaptation of Large Language Models (Hu et al., 2021). Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/d7b97d1b93d8d8bc.json.</description>
      <category>lora, fine-tuning, foundational, hu, 2021, microsoft</category>
    </item>
    <item>
      <title>Direct Preference Optimization (DPO) introduced in paper: Direct Preference Optimization: Your Language Model is Secretly a Reward Model (Rafailov et al., 2023).</title>
      <link>https://sourcescore.org/claims/a3e691683a4577af/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/a3e691683a4577af/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Direct Preference Optimization (DPO) introduced in paper: Direct Preference Optimization: Your Language Model is Secretly a Reward Model (Rafailov et al., 2023). Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/a3e691683a4577af.json.</description>
      <category>dpo, alignment, foundational, rafailov, 2023, nips, stanford</category>
    </item>
    <item>
      <title>Constitutional AI (CAI) introduced in paper: Constitutional AI: Harmlessness from AI Feedback (Bai et al., 2022).</title>
      <link>https://sourcescore.org/claims/ba1eb83c14795107/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/ba1eb83c14795107/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Constitutional AI (CAI) introduced in paper: Constitutional AI: Harmlessness from AI Feedback (Bai et al., 2022). Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/ba1eb83c14795107.json.</description>
      <category>constitutional-ai, alignment, anthropic, 2022, bai</category>
    </item>
    <item>
      <title>InstructGPT methodology introduced in paper: Training language models to follow instructions with human feedback (Ouyang et al., 2022).</title>
      <link>https://sourcescore.org/claims/5da8f8dffc038b8e/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/5da8f8dffc038b8e/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>InstructGPT methodology introduced in paper: Training language models to follow instructions with human feedback (Ouyang et al., 2022). Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/5da8f8dffc038b8e.json.</description>
      <category>instructgpt, alignment, openai, 2022, ouyang, rlhf</category>
    </item>
    <item>
      <title>FlashAttention introduced in paper: FlashAttention: Fast and Memory-Efficient Exact Attention with IO-Awareness (Dao et al., 2022).</title>
      <link>https://sourcescore.org/claims/e120182d1e01ea2b/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/e120182d1e01ea2b/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>FlashAttention introduced in paper: FlashAttention: Fast and Memory-Efficient Exact Attention with IO-Awareness (Dao et al., 2022). Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/e120182d1e01ea2b.json.</description>
      <category>flash-attention, performance, dao, 2022, stanford</category>
    </item>
    <item>
      <title>CLIP (Contrastive Language-Image Pretraining) introduced in paper: Learning Transferable Visual Models From Natural Language Supervision (Radford et al., 2021).</title>
      <link>https://sourcescore.org/claims/85a3ca745eaf4ee0/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/85a3ca745eaf4ee0/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>CLIP (Contrastive Language-Image Pretraining) introduced in paper: Learning Transferable Visual Models From Natural Language Supervision (Radford et al., 2021). Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/85a3ca745eaf4ee0.json.</description>
      <category>clip, multimodal, vision, radford, 2021, openai</category>
    </item>
    <item>
      <title>ChatGPT released on: 2022-11-30.</title>
      <link>https://sourcescore.org/claims/8d653880c519a8ef/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/8d653880c519a8ef/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>ChatGPT released on: 2022-11-30. Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/8d653880c519a8ef.json.</description>
      <category>chatgpt, openai, release, 2022, gpt-3.5</category>
    </item>
    <item>
      <title>GPT-4 released on: 2023-03-14.</title>
      <link>https://sourcescore.org/claims/09eea8fb1a8ccebf/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/09eea8fb1a8ccebf/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>GPT-4 released on: 2023-03-14. Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/09eea8fb1a8ccebf.json.</description>
      <category>gpt-4, openai, release, 2023</category>
    </item>
    <item>
      <title>GPT-4 Turbo context window tokens: 128000.</title>
      <link>https://sourcescore.org/claims/26314e3164e18b24/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/26314e3164e18b24/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>GPT-4 Turbo context window tokens: 128000. Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/26314e3164e18b24.json.</description>
      <category>gpt-4-turbo, context, openai, 128k, devday</category>
    </item>
    <item>
      <title>GPT-4o released on: 2024-05-13.</title>
      <link>https://sourcescore.org/claims/bd065b91ca6e880b/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/bd065b91ca6e880b/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>GPT-4o released on: 2024-05-13. Verified 2026-05-16. Confidence 100%. 1 primary source. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/bd065b91ca6e880b.json.</description>
      <category>gpt-4o, openai, release, 2024, multimodal</category>
    </item>
    <item>
      <title>Claude 3.5 Sonnet released on: 2024-06-20.</title>
      <link>https://sourcescore.org/claims/9fc5fb203abbf250/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/9fc5fb203abbf250/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Claude 3.5 Sonnet released on: 2024-06-20. Verified 2026-05-16. Confidence 100%. 1 primary source. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/9fc5fb203abbf250.json.</description>
      <category>claude-3.5, sonnet, anthropic, release, 2024</category>
    </item>
    <item>
      <title>Claude 3 Opus context window tokens: 200000.</title>
      <link>https://sourcescore.org/claims/565df27fc8b75ef0/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/565df27fc8b75ef0/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Claude 3 Opus context window tokens: 200000. Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/565df27fc8b75ef0.json.</description>
      <category>claude-3, opus, anthropic, context, 200k</category>
    </item>
    <item>
      <title>Llama 2 released on: 2023-07-18.</title>
      <link>https://sourcescore.org/claims/34dd74941bc7bd48/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/34dd74941bc7bd48/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Llama 2 released on: 2023-07-18. Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/34dd74941bc7bd48.json.</description>
      <category>llama-2, meta, release, 2023, open-weights</category>
    </item>
    <item>
      <title>Llama 3.1 released on: 2024-07-23.</title>
      <link>https://sourcescore.org/claims/a55484ab8b4bdf4e/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/a55484ab8b4bdf4e/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Llama 3.1 released on: 2024-07-23. Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/a55484ab8b4bdf4e.json.</description>
      <category>llama-3.1, meta, release, 2024, open-weights, 405b</category>
    </item>
    <item>
      <title>Llama 3.1 405B parameter count: 405000000000.</title>
      <link>https://sourcescore.org/claims/282b0523eefd9afd/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/282b0523eefd9afd/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Llama 3.1 405B parameter count: 405000000000. Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/282b0523eefd9afd.json.</description>
      <category>llama-3.1, parameters, 405b, meta</category>
    </item>
    <item>
      <title>Mixtral 8x7B released on: 2023-12-11.</title>
      <link>https://sourcescore.org/claims/410aec4f418f2b11/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/410aec4f418f2b11/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Mixtral 8x7B released on: 2023-12-11. Verified 2026-05-16. Confidence 95%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/410aec4f418f2b11.json.</description>
      <category>mixtral, mistral, release, 2023, moe, sparse</category>
    </item>
    <item>
      <title>Mixtral 8x7B architecture: Sparse Mixture-of-Experts (8 experts × 7B params, 2 experts routed per token).</title>
      <link>https://sourcescore.org/claims/ad79b14fafb362cd/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/ad79b14fafb362cd/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Mixtral 8x7B architecture: Sparse Mixture-of-Experts (8 experts × 7B params, 2 experts routed per token). Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/ad79b14fafb362cd.json.</description>
      <category>mixtral, moe, architecture, mistral</category>
    </item>
    <item>
      <title>Gemini Pro released on: 2023-12-06.</title>
      <link>https://sourcescore.org/claims/e2a6019bd2ce5c97/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/e2a6019bd2ce5c97/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Gemini Pro released on: 2023-12-06. Verified 2026-05-16. Confidence 100%. 1 primary source. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/e2a6019bd2ce5c97.json.</description>
      <category>gemini, google, release, 2023</category>
    </item>
    <item>
      <title>Whisper released on: 2022-09-21.</title>
      <link>https://sourcescore.org/claims/a3ebbaed14bd83d0/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/a3ebbaed14bd83d0/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Whisper released on: 2022-09-21. Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/a3ebbaed14bd83d0.json.</description>
      <category>whisper, openai, speech, asr, release, 2022</category>
    </item>
    <item>
      <title>Anthropic founded in: 2021.</title>
      <link>https://sourcescore.org/claims/dc53356a0a39c8de/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/dc53356a0a39c8de/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Anthropic founded in: 2021. Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/dc53356a0a39c8de.json.</description>
      <category>anthropic, company, founded, 2021</category>
    </item>
    <item>
      <title>OpenAI founded in: 2015.</title>
      <link>https://sourcescore.org/claims/64a2379efebe62ee/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/64a2379efebe62ee/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>OpenAI founded in: 2015. Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/64a2379efebe62ee.json.</description>
      <category>openai, company, founded, 2015</category>
    </item>
    <item>
      <title>Mistral AI founded in: 2023.</title>
      <link>https://sourcescore.org/claims/db8e97f3583db317/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/db8e97f3583db317/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Mistral AI founded in: 2023. Verified 2026-05-16. Confidence 100%. 1 primary source. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/db8e97f3583db317.json.</description>
      <category>mistral, company, founded, 2023</category>
    </item>
    <item>
      <title>Hugging Face founded in: 2016.</title>
      <link>https://sourcescore.org/claims/0e28201946131976/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/0e28201946131976/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Hugging Face founded in: 2016. Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/0e28201946131976.json.</description>
      <category>hugging-face, company, founded, 2016</category>
    </item>
    <item>
      <title>BERT (Bidirectional Encoder Representations from Transformers) introduced in paper: BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding (Devlin et al., 2018).</title>
      <link>https://sourcescore.org/claims/4c1ee70007dc89c1/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/4c1ee70007dc89c1/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>BERT (Bidirectional Encoder Representations from Transformers) introduced in paper: BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding (Devlin et al., 2018). Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/4c1ee70007dc89c1.json.</description>
      <category>bert, foundational, devlin, 2018, google, nlp</category>
    </item>
    <item>
      <title>GPT-2 introduced in paper: Language Models are Unsupervised Multitask Learners (Radford et al., 2019).</title>
      <link>https://sourcescore.org/claims/859551dc078c46f8/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/859551dc078c46f8/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>GPT-2 introduced in paper: Language Models are Unsupervised Multitask Learners (Radford et al., 2019). Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/859551dc078c46f8.json.</description>
      <category>gpt-2, foundational, openai, radford, 2019</category>
    </item>
    <item>
      <title>ResNet (Residual Networks) introduced in paper: Deep Residual Learning for Image Recognition (He et al., 2015).</title>
      <link>https://sourcescore.org/claims/4f55f77c4bfb316e/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/4f55f77c4bfb316e/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>ResNet (Residual Networks) introduced in paper: Deep Residual Learning for Image Recognition (He et al., 2015). Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/4f55f77c4bfb316e.json.</description>
      <category>resnet, foundational, vision, he, 2015, microsoft, cvpr</category>
    </item>
    <item>
      <title>T5 (Text-to-Text Transfer Transformer) introduced in paper: Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer (Raffel et al., 2019).</title>
      <link>https://sourcescore.org/claims/ef28341c3b308737/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/ef28341c3b308737/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>T5 (Text-to-Text Transfer Transformer) introduced in paper: Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer (Raffel et al., 2019). Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/ef28341c3b308737.json.</description>
      <category>t5, foundational, transfer-learning, raffel, 2019, google</category>
    </item>
    <item>
      <title>Sparsely-Gated Mixture-of-Experts (MoE) introduced in paper: Outrageously Large Neural Networks: The Sparsely-Gated Mixture-of-Experts Layer (Shazeer et al., 2017).</title>
      <link>https://sourcescore.org/claims/2d6d7f61f1db6493/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/2d6d7f61f1db6493/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Sparsely-Gated Mixture-of-Experts (MoE) introduced in paper: Outrageously Large Neural Networks: The Sparsely-Gated Mixture-of-Experts Layer (Shazeer et al., 2017). Verified 2026-05-16. Confidence 100%. 1 primary source. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/2d6d7f61f1db6493.json.</description>
      <category>moe, foundational, shazeer, 2017, google</category>
    </item>
    <item>
      <title>Switch Transformer introduced in paper: Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity (Fedus et al., 2021).</title>
      <link>https://sourcescore.org/claims/3d9c14b9379038c9/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/3d9c14b9379038c9/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Switch Transformer introduced in paper: Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity (Fedus et al., 2021). Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/3d9c14b9379038c9.json.</description>
      <category>switch-transformer, moe, foundational, fedus, 2021, google</category>
    </item>
    <item>
      <title>Chinchilla scaling laws introduced in paper: Training Compute-Optimal Large Language Models (Hoffmann et al., 2022).</title>
      <link>https://sourcescore.org/claims/8befcae6bce01a95/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/8befcae6bce01a95/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Chinchilla scaling laws introduced in paper: Training Compute-Optimal Large Language Models (Hoffmann et al., 2022). Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/8befcae6bce01a95.json.</description>
      <category>chinchilla, scaling-laws, foundational, hoffmann, 2022, deepmind, nips</category>
    </item>
    <item>
      <title>Proximal Policy Optimization (PPO) introduced in paper: Proximal Policy Optimization Algorithms (Schulman et al., 2017).</title>
      <link>https://sourcescore.org/claims/00f224e1ccc158ef/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/00f224e1ccc158ef/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Proximal Policy Optimization (PPO) introduced in paper: Proximal Policy Optimization Algorithms (Schulman et al., 2017). Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/00f224e1ccc158ef.json.</description>
      <category>ppo, reinforcement-learning, foundational, schulman, 2017, openai, rlhf</category>
    </item>
    <item>
      <title>Mamba state-space model introduced in paper: Mamba: Linear-Time Sequence Modeling with Selective State Spaces (Gu, Dao, 2023).</title>
      <link>https://sourcescore.org/claims/3518f8aa40cb0d36/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/3518f8aa40cb0d36/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Mamba state-space model introduced in paper: Mamba: Linear-Time Sequence Modeling with Selective State Spaces (Gu, Dao, 2023). Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/3518f8aa40cb0d36.json.</description>
      <category>mamba, state-space, foundational, gu, dao, 2023</category>
    </item>
    <item>
      <title>Chain-of-Thought prompting introduced in paper: Chain-of-Thought Prompting Elicits Reasoning in Large Language Models (Wei et al., 2022).</title>
      <link>https://sourcescore.org/claims/3af924da138ff84c/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/3af924da138ff84c/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Chain-of-Thought prompting introduced in paper: Chain-of-Thought Prompting Elicits Reasoning in Large Language Models (Wei et al., 2022). Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/3af924da138ff84c.json.</description>
      <category>chain-of-thought, cot, prompting, foundational, wei, 2022, google, nips</category>
    </item>
    <item>
      <title>Adam optimizer introduced in paper: Adam: A Method for Stochastic Optimization (Kingma, Ba, 2014).</title>
      <link>https://sourcescore.org/claims/dffbe905003cc581/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/dffbe905003cc581/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Adam optimizer introduced in paper: Adam: A Method for Stochastic Optimization (Kingma, Ba, 2014). Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/dffbe905003cc581.json.</description>
      <category>adam, optimizer, foundational, kingma, 2014, iclr</category>
    </item>
    <item>
      <title>AlexNet introduced in paper: ImageNet Classification with Deep Convolutional Neural Networks (Krizhevsky, Sutskever, Hinton, 2012).</title>
      <link>https://sourcescore.org/claims/98b6e774be89d967/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/98b6e774be89d967/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>AlexNet introduced in paper: ImageNet Classification with Deep Convolutional Neural Networks (Krizhevsky, Sutskever, Hinton, 2012). Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/98b6e774be89d967.json.</description>
      <category>alexnet, foundational, vision, krizhevsky, hinton, 2012, nips, imagenet</category>
    </item>
    <item>
      <title>Llama 1 released on: 2023-02-24.</title>
      <link>https://sourcescore.org/claims/5ce2e90fafdca0c2/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/5ce2e90fafdca0c2/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Llama 1 released on: 2023-02-24. Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/5ce2e90fafdca0c2.json.</description>
      <category>llama-1, meta, release, 2023</category>
    </item>
    <item>
      <title>Llama 3 released on: 2024-04-18.</title>
      <link>https://sourcescore.org/claims/5f599876b3dd19b3/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/5f599876b3dd19b3/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Llama 3 released on: 2024-04-18. Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/5f599876b3dd19b3.json.</description>
      <category>llama-3, meta, release, 2024, open-weights</category>
    </item>
    <item>
      <title>Llama 3 70B parameter count: 70000000000.</title>
      <link>https://sourcescore.org/claims/e4581034693f4584/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/e4581034693f4584/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Llama 3 70B parameter count: 70000000000. Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/e4581034693f4584.json.</description>
      <category>llama-3, parameters, 70b, meta</category>
    </item>
    <item>
      <title>Claude 3 family (Opus, Sonnet, Haiku) released on: 2024-03-04.</title>
      <link>https://sourcescore.org/claims/99340a18716a44fb/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/99340a18716a44fb/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Claude 3 family (Opus, Sonnet, Haiku) released on: 2024-03-04. Verified 2026-05-16. Confidence 100%. 1 primary source. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/99340a18716a44fb.json.</description>
      <category>claude-3, opus, sonnet, haiku, anthropic, release, 2024</category>
    </item>
    <item>
      <title>Gemini 1.5 Pro announced on: 2024-02-15.</title>
      <link>https://sourcescore.org/claims/b55da789dba1c76b/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/b55da789dba1c76b/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Gemini 1.5 Pro announced on: 2024-02-15. Verified 2026-05-16. Confidence 100%. 1 primary source. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/b55da789dba1c76b.json.</description>
      <category>gemini-1.5, google, announcement, 2024</category>
    </item>
    <item>
      <title>Mistral 7B released on: 2023-09-27.</title>
      <link>https://sourcescore.org/claims/2d5a5dc30e7f6f02/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/2d5a5dc30e7f6f02/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Mistral 7B released on: 2023-09-27. Verified 2026-05-16. Confidence 100%. 3 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/2d5a5dc30e7f6f02.json.</description>
      <category>mistral-7b, mistral, release, 2023, open-weights</category>
    </item>
    <item>
      <title>Stable Diffusion 1.0 released on: 2022-08-22.</title>
      <link>https://sourcescore.org/claims/f6f333228f224df2/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/f6f333228f224df2/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Stable Diffusion 1.0 released on: 2022-08-22. Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/f6f333228f224df2.json.</description>
      <category>stable-diffusion, stability-ai, release, 2022, image-generation</category>
    </item>
    <item>
      <title>DALL-E 3 announced on: 2023-09-20.</title>
      <link>https://sourcescore.org/claims/ce5275632f16109a/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/ce5275632f16109a/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>DALL-E 3 announced on: 2023-09-20. Verified 2026-05-16. Confidence 100%. 1 primary source. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/ce5275632f16109a.json.</description>
      <category>dall-e-3, openai, image-generation, 2023</category>
    </item>
    <item>
      <title>Sora (OpenAI video generation model) announced on: 2024-02-15.</title>
      <link>https://sourcescore.org/claims/5697b2a15693e6c0/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/5697b2a15693e6c0/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Sora (OpenAI video generation model) announced on: 2024-02-15. Verified 2026-05-16. Confidence 100%. 1 primary source. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/5697b2a15693e6c0.json.</description>
      <category>sora, openai, video-generation, 2024</category>
    </item>
    <item>
      <title>ChatGPT Plus launched on: 2023-02-01.</title>
      <link>https://sourcescore.org/claims/461dd4dbccd9038d/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/461dd4dbccd9038d/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>ChatGPT Plus launched on: 2023-02-01. Verified 2026-05-16. Confidence 100%. 1 primary source. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/461dd4dbccd9038d.json.</description>
      <category>chatgpt-plus, openai, subscription, launch, 2023</category>
    </item>
    <item>
      <title>GPT-3 parameter count: 175000000000.</title>
      <link>https://sourcescore.org/claims/1ca2cc2864dfb376/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/1ca2cc2864dfb376/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>GPT-3 parameter count: 175000000000. Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/1ca2cc2864dfb376.json.</description>
      <category>gpt-3, openai, parameters, 175b, brown, 2020</category>
    </item>
    <item>
      <title>Stability AI founded in: 2019.</title>
      <link>https://sourcescore.org/claims/d23bd1687a3ea44a/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/d23bd1687a3ea44a/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Stability AI founded in: 2019. Verified 2026-05-16. Confidence 95%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/d23bd1687a3ea44a.json.</description>
      <category>stability-ai, company, founded, 2019</category>
    </item>
    <item>
      <title>DeepMind acquired by google on: 2014-01-26.</title>
      <link>https://sourcescore.org/claims/25a8f486a0947790/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/25a8f486a0947790/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>DeepMind acquired by google on: 2014-01-26. Verified 2026-05-16. Confidence 95%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/25a8f486a0947790.json.</description>
      <category>deepmind, google, acquisition, 2014, company</category>
    </item>
    <item>
      <title>ImageNet dataset introduced in paper: ImageNet: A Large-Scale Hierarchical Image Database (Deng et al., 2009).</title>
      <link>https://sourcescore.org/claims/045e628def62181d/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/045e628def62181d/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>ImageNet dataset introduced in paper: ImageNet: A Large-Scale Hierarchical Image Database (Deng et al., 2009). Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/045e628def62181d.json.</description>
      <category>imagenet, dataset, vision, foundational, fei-fei-li, 2009, cvpr</category>
    </item>
    <item>
      <title>Vision Transformer (ViT) introduced in paper: An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale (Dosovitskiy et al., 2020).</title>
      <link>https://sourcescore.org/claims/d3681b0981e0b700/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/d3681b0981e0b700/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Vision Transformer (ViT) introduced in paper: An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale (Dosovitskiy et al., 2020). Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/d3681b0981e0b700.json.</description>
      <category>vit, vision-transformer, foundational, dosovitskiy, 2020, google, iclr</category>
    </item>
    <item>
      <title>AlphaFold 2 published in paper: Highly accurate protein structure prediction with AlphaFold (Jumper et al., 2021).</title>
      <link>https://sourcescore.org/claims/477bc19212addf9a/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/477bc19212addf9a/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>AlphaFold 2 published in paper: Highly accurate protein structure prediction with AlphaFold (Jumper et al., 2021). Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/477bc19212addf9a.json.</description>
      <category>alphafold, deepmind, biology, 2021, nature, protein</category>
    </item>
    <item>
      <title>Generative Adversarial Networks (GANs) introduced in paper: Generative Adversarial Networks (Goodfellow et al., 2014).</title>
      <link>https://sourcescore.org/claims/5b0c0612bd9e55b0/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/5b0c0612bd9e55b0/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Generative Adversarial Networks (GANs) introduced in paper: Generative Adversarial Networks (Goodfellow et al., 2014). Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/5b0c0612bd9e55b0.json.</description>
      <category>gan, foundational, goodfellow, 2014, nips, generative</category>
    </item>
    <item>
      <title>Variational Autoencoder (VAE) introduced in paper: Auto-Encoding Variational Bayes (Kingma, Welling, 2013).</title>
      <link>https://sourcescore.org/claims/62789e45973ab631/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/62789e45973ab631/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Variational Autoencoder (VAE) introduced in paper: Auto-Encoding Variational Bayes (Kingma, Welling, 2013). Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/62789e45973ab631.json.</description>
      <category>vae, foundational, kingma, welling, 2013, iclr, generative</category>
    </item>
    <item>
      <title>Denoising Diffusion Probabilistic Models (DDPM) introduced in paper: Denoising Diffusion Probabilistic Models (Ho, Jain, Abbeel, 2020).</title>
      <link>https://sourcescore.org/claims/e700f81fff6f38c7/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/e700f81fff6f38c7/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Denoising Diffusion Probabilistic Models (DDPM) introduced in paper: Denoising Diffusion Probabilistic Models (Ho, Jain, Abbeel, 2020). Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/e700f81fff6f38c7.json.</description>
      <category>ddpm, diffusion, foundational, ho, 2020, nips, image-generation</category>
    </item>
    <item>
      <title>Word2Vec introduced in paper: Efficient Estimation of Word Representations in Vector Space (Mikolov et al., 2013).</title>
      <link>https://sourcescore.org/claims/4978f76d228a3db1/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/4978f76d228a3db1/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Word2Vec introduced in paper: Efficient Estimation of Word Representations in Vector Space (Mikolov et al., 2013). Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/4978f76d228a3db1.json.</description>
      <category>word2vec, embeddings, foundational, mikolov, 2013, google, nlp</category>
    </item>
    <item>
      <title>Byte-Pair Encoding (BPE) for Neural Machine Translation introduced in paper: Neural Machine Translation of Rare Words with Subword Units (Sennrich et al., 2015).</title>
      <link>https://sourcescore.org/claims/e942c93d70a4dab2/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/e942c93d70a4dab2/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Byte-Pair Encoding (BPE) for Neural Machine Translation introduced in paper: Neural Machine Translation of Rare Words with Subword Units (Sennrich et al., 2015). Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/e942c93d70a4dab2.json.</description>
      <category>bpe, tokenization, foundational, sennrich, 2015, acl, nmt</category>
    </item>
    <item>
      <title>MMLU benchmark introduced in paper: Measuring Massive Multitask Language Understanding (Hendrycks et al., 2020).</title>
      <link>https://sourcescore.org/claims/428d754e7c651be6/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/428d754e7c651be6/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>MMLU benchmark introduced in paper: Measuring Massive Multitask Language Understanding (Hendrycks et al., 2020). Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/428d754e7c651be6.json.</description>
      <category>mmlu, benchmark, hendrycks, 2020, iclr, evaluation</category>
    </item>
    <item>
      <title>HumanEval benchmark introduced in paper: Evaluating Large Language Models Trained on Code (Chen et al., 2021).</title>
      <link>https://sourcescore.org/claims/71ec42731d2c9e0c/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/71ec42731d2c9e0c/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>HumanEval benchmark introduced in paper: Evaluating Large Language Models Trained on Code (Chen et al., 2021). Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/71ec42731d2c9e0c.json.</description>
      <category>humaneval, benchmark, codex, openai, chen, 2021, code-generation</category>
    </item>
    <item>
      <title>Llama 2 70B parameter count: 70000000000.</title>
      <link>https://sourcescore.org/claims/a4fc4391a27f0500/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/a4fc4391a27f0500/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Llama 2 70B parameter count: 70000000000. Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/a4fc4391a27f0500.json.</description>
      <category>llama-2, parameters, 70b, meta</category>
    </item>
    <item>
      <title>Llama 3 8B parameter count: 8000000000.</title>
      <link>https://sourcescore.org/claims/8455a8b9d44fecbe/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/8455a8b9d44fecbe/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Llama 3 8B parameter count: 8000000000. Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/8455a8b9d44fecbe.json.</description>
      <category>llama-3, parameters, 8b, meta</category>
    </item>
    <item>
      <title>GPT-4o mini released on: 2024-07-18.</title>
      <link>https://sourcescore.org/claims/1a4cf45e8b967089/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/1a4cf45e8b967089/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>GPT-4o mini released on: 2024-07-18. Verified 2026-05-16. Confidence 100%. 1 primary source. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/1a4cf45e8b967089.json.</description>
      <category>gpt-4o-mini, openai, release, 2024</category>
    </item>
    <item>
      <title>Claude 3.5 Haiku released on: 2024-11-04.</title>
      <link>https://sourcescore.org/claims/cdc82d4d6c880ba3/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/cdc82d4d6c880ba3/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Claude 3.5 Haiku released on: 2024-11-04. Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/cdc82d4d6c880ba3.json.</description>
      <category>claude-3.5-haiku, anthropic, release, 2024</category>
    </item>
    <item>
      <title>Phi-3 (Microsoft small language model family) released on: 2024-04-23.</title>
      <link>https://sourcescore.org/claims/2d20c4a92620ba46/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/2d20c4a92620ba46/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Phi-3 (Microsoft small language model family) released on: 2024-04-23. Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/2d20c4a92620ba46.json.</description>
      <category>phi-3, microsoft, release, 2024, slm</category>
    </item>
    <item>
      <title>DeepSeek V3 released on: 2024-12-26.</title>
      <link>https://sourcescore.org/claims/6aa7e9c75d084617/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/6aa7e9c75d084617/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>DeepSeek V3 released on: 2024-12-26. Verified 2026-05-16. Confidence 95%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/6aa7e9c75d084617.json.</description>
      <category>deepseek-v3, deepseek, release, 2024, open-weights</category>
    </item>
    <item>
      <title>Llama 3.2 (multimodal release including 11B and 90B vision models) released on: 2024-09-25.</title>
      <link>https://sourcescore.org/claims/e27816c692a28ce9/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/e27816c692a28ce9/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Llama 3.2 (multimodal release including 11B and 90B vision models) released on: 2024-09-25. Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/e27816c692a28ce9.json.</description>
      <category>llama-3.2, meta, release, 2024, multimodal, vision</category>
    </item>
    <item>
      <title>Llama 3.3 70B released on: 2024-12-06.</title>
      <link>https://sourcescore.org/claims/97c04944b4c6345a/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/97c04944b4c6345a/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Llama 3.3 70B released on: 2024-12-06. Verified 2026-05-16. Confidence 100%. 1 primary source. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/97c04944b4c6345a.json.</description>
      <category>llama-3.3, meta, release, 2024, open-weights</category>
    </item>
    <item>
      <title>Gemini 2.0 announced on: 2024-12-11.</title>
      <link>https://sourcescore.org/claims/bc1d6baa80070e80/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/bc1d6baa80070e80/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Gemini 2.0 announced on: 2024-12-11. Verified 2026-05-16. Confidence 100%. 1 primary source. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/bc1d6baa80070e80.json.</description>
      <category>gemini-2.0, google, deepmind, announcement, 2024</category>
    </item>
    <item>
      <title>Grok (xAI) announced on: 2023-11-04.</title>
      <link>https://sourcescore.org/claims/930c3c71363acee5/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/930c3c71363acee5/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Grok (xAI) announced on: 2023-11-04. Verified 2026-05-16. Confidence 95%. 1 primary source. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/930c3c71363acee5.json.</description>
      <category>grok, xai, announcement, 2023</category>
    </item>
    <item>
      <title>GPT-4o context window tokens: 128000.</title>
      <link>https://sourcescore.org/claims/118226d52c6f7491/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/118226d52c6f7491/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>GPT-4o context window tokens: 128000. Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/118226d52c6f7491.json.</description>
      <category>gpt-4o, context, openai, 128k</category>
    </item>
    <item>
      <title>Sora (public release to ChatGPT Plus subscribers) released on: 2024-12-09.</title>
      <link>https://sourcescore.org/claims/409b828ed9f8410b/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/409b828ed9f8410b/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Sora (public release to ChatGPT Plus subscribers) released on: 2024-12-09. Verified 2026-05-16. Confidence 100%. 1 primary source. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/409b828ed9f8410b.json.</description>
      <category>sora, openai, release, 2024, video-generation</category>
    </item>
    <item>
      <title>Cohere founded in: 2019.</title>
      <link>https://sourcescore.org/claims/cf4cbf71dc7dd19a/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/cf4cbf71dc7dd19a/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Cohere founded in: 2019. Verified 2026-05-16. Confidence 95%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/cf4cbf71dc7dd19a.json.</description>
      <category>cohere, company, founded, 2019</category>
    </item>
    <item>
      <title>xAI founded in: 2023.</title>
      <link>https://sourcescore.org/claims/e5084c3f7023114b/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/e5084c3f7023114b/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>xAI founded in: 2023. Verified 2026-05-16. Confidence 95%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/e5084c3f7023114b.json.</description>
      <category>xai, company, founded, 2023, musk</category>
    </item>
    <item>
      <title>Common Crawl founded in: 2007.</title>
      <link>https://sourcescore.org/claims/4a2689e6230ef2e1/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/4a2689e6230ef2e1/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Common Crawl founded in: 2007. Verified 2026-05-16. Confidence 95%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/4a2689e6230ef2e1.json.</description>
      <category>common-crawl, dataset, founded, 2007</category>
    </item>
    <item>
      <title>ReAct (Reasoning + Acting) introduced in paper: ReAct: Synergizing Reasoning and Acting in Language Models (Yao et al., 2022).</title>
      <link>https://sourcescore.org/claims/fceea64fa7d04d3a/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/fceea64fa7d04d3a/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>ReAct (Reasoning + Acting) introduced in paper: ReAct: Synergizing Reasoning and Acting in Language Models (Yao et al., 2022). Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/fceea64fa7d04d3a.json.</description>
      <category>react, reasoning, agents, tool-use, foundational, 2022, yao</category>
    </item>
    <item>
      <title>LoRA (Low-Rank Adaptation) introduced in paper: LoRA: Low-Rank Adaptation of Large Language Models (Hu et al., 2021).</title>
      <link>https://sourcescore.org/claims/f191b2876790dc6e/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/f191b2876790dc6e/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>LoRA (Low-Rank Adaptation) introduced in paper: LoRA: Low-Rank Adaptation of Large Language Models (Hu et al., 2021). Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/f191b2876790dc6e.json.</description>
      <category>lora, peft, fine-tuning, foundational, 2021, microsoft</category>
    </item>
    <item>
      <title>QLoRA introduced in paper: QLoRA: Efficient Finetuning of Quantized LLMs (Dettmers et al., 2023).</title>
      <link>https://sourcescore.org/claims/767cbe41c961be1a/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/767cbe41c961be1a/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>QLoRA introduced in paper: QLoRA: Efficient Finetuning of Quantized LLMs (Dettmers et al., 2023). Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/767cbe41c961be1a.json.</description>
      <category>qlora, quantization, peft, fine-tuning, foundational, 2023</category>
    </item>
    <item>
      <title>Rotary Position Embedding (RoPE) introduced in paper: RoFormer: Enhanced Transformer with Rotary Position Embedding (Su et al., 2021).</title>
      <link>https://sourcescore.org/claims/f8d64457ba9fd35b/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/f8d64457ba9fd35b/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Rotary Position Embedding (RoPE) introduced in paper: RoFormer: Enhanced Transformer with Rotary Position Embedding (Su et al., 2021). Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/f8d64457ba9fd35b.json.</description>
      <category>rope, position-embedding, transformer, foundational, 2021</category>
    </item>
    <item>
      <title>Byte-Pair Encoding (BPE) for NMT introduced in paper: Neural Machine Translation of Rare Words with Subword Units (Sennrich et al., 2015).</title>
      <link>https://sourcescore.org/claims/aede848e23c8de8e/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/aede848e23c8de8e/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Byte-Pair Encoding (BPE) for NMT introduced in paper: Neural Machine Translation of Rare Words with Subword Units (Sennrich et al., 2015). Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/aede848e23c8de8e.json.</description>
      <category>bpe, tokenization, subword, foundational, 2015, acl</category>
    </item>
    <item>
      <title>SentencePiece tokenizer introduced in paper: SentencePiece: A simple and language independent subword tokenizer and detokenizer for Neural Text Processing (Kudo &amp; Richardson, 2018).</title>
      <link>https://sourcescore.org/claims/0d47bb8eb637a2e4/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/0d47bb8eb637a2e4/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>SentencePiece tokenizer introduced in paper: SentencePiece: A simple and language independent subword tokenizer and detokenizer for Neural Text Processing (Kudo &amp; Richardson, 2018). Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/0d47bb8eb637a2e4.json.</description>
      <category>sentencepiece, tokenization, google, foundational, 2018</category>
    </item>
    <item>
      <title>C4 (Colossal Clean Crawled Corpus) introduced in paper: Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer (Raffel et al., 2019).</title>
      <link>https://sourcescore.org/claims/0d24c97977ebd744/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/0d24c97977ebd744/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>C4 (Colossal Clean Crawled Corpus) introduced in paper: Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer (Raffel et al., 2019). Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/0d24c97977ebd744.json.</description>
      <category>c4, dataset, pretraining, google, 2019</category>
    </item>
    <item>
      <title>The Pile dataset released on: 2020-12-31.</title>
      <link>https://sourcescore.org/claims/4aef1422b96df26c/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/4aef1422b96df26c/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>The Pile dataset released on: 2020-12-31. Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/4aef1422b96df26c.json.</description>
      <category>the-pile, dataset, pretraining, eleutherai, 2020</category>
    </item>
    <item>
      <title>RedPajama dataset released on: 2023-04-17.</title>
      <link>https://sourcescore.org/claims/ea8b7be3a49101be/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/ea8b7be3a49101be/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>RedPajama dataset released on: 2023-04-17. Verified 2026-05-16. Confidence 95%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/ea8b7be3a49101be.json.</description>
      <category>redpajama, dataset, pretraining, together, 2023, open-source</category>
    </item>
    <item>
      <title>CLIP introduced in paper: Learning Transferable Visual Models From Natural Language Supervision (Radford et al., 2021).</title>
      <link>https://sourcescore.org/claims/bcdef949cc6d3644/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/bcdef949cc6d3644/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>CLIP introduced in paper: Learning Transferable Visual Models From Natural Language Supervision (Radford et al., 2021). Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/bcdef949cc6d3644.json.</description>
      <category>clip, vision-language, multimodal, foundational, 2021, openai</category>
    </item>
    <item>
      <title>DALL·E 2 released on: 2022-04-06.</title>
      <link>https://sourcescore.org/claims/0b0e64476bd25bd6/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/0b0e64476bd25bd6/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>DALL·E 2 released on: 2022-04-06. Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/0b0e64476bd25bd6.json.</description>
      <category>dall-e, dall-e-2, image-generation, multimodal, openai, released_on, 2022</category>
    </item>
    <item>
      <title>Stable Diffusion 1.x released on: 2022-08-22.</title>
      <link>https://sourcescore.org/claims/79a7e980e59680bc/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/79a7e980e59680bc/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Stable Diffusion 1.x released on: 2022-08-22. Verified 2026-05-16. Confidence 100%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/79a7e980e59680bc.json.</description>
      <category>stable-diffusion, image-generation, stability-ai, released_on, 2022</category>
    </item>
    <item>
      <title>EleutherAI founded in: 2020.</title>
      <link>https://sourcescore.org/claims/f018fec775a8e941/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/f018fec775a8e941/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>EleutherAI founded in: 2020. Verified 2026-05-16. Confidence 95%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/f018fec775a8e941.json.</description>
      <category>eleutherai, company, founded, 2020, open-source</category>
    </item>
    <item>
      <title>Together AI founded in: 2022.</title>
      <link>https://sourcescore.org/claims/4d05fa0726841c32/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/4d05fa0726841c32/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>Together AI founded in: 2022. Verified 2026-05-16. Confidence 90%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/4d05fa0726841c32.json.</description>
      <category>together-ai, company, founded, 2022</category>
    </item>
    <item>
      <title>AI21 Labs founded in: 2017.</title>
      <link>https://sourcescore.org/claims/c7c1ee139886cafe/</link>
      <guid isPermaLink="true">https://sourcescore.org/claims/c7c1ee139886cafe/</guid>
      <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
      <description>AI21 Labs founded in: 2017. Verified 2026-05-16. Confidence 95%. 2 primary sources. HMAC-SHA256 signed envelope at https://sourcescore.org/api/v1/claims/c7c1ee139886cafe.json.</description>
      <category>ai21, ai21-labs, company, founded, 2017, israel</category>
    </item>
  </channel>
</rss>
