official
· 154 artifacts, sorted by favorites. · open in search — combine tags, sort, filter by date →
- Worse Than MechaHitler ♥0
- Opus 4.7 Part 3: Model Welfare ♥0
- Opus 4.7 Part 2: Capabilities and Reactions ♥0
- Opus 4.7 Part 1: The Model Card ♥0
- Claude Opus 4.5: Model Card, Alignment and Safety ♥0
- Claude Opus 4.5 Is The Best Model Available ♥0
- The Once And Future Fable #2 ♥0
- On GPT-4.5 ♥0
- On DeepSeek's r1 ♥0
- On Claude 3.5 Sonnet ♥0
- On Claude 3.0 ♥0
- o3 Will Use Its Tools For You ♥0
- Zvi: o1 Turns Pro ♥0
- Zvi: The o1 System Card Is Not About o1 ♥0
- No, Grok, No ♥0
- Kimi K2 ♥0
- I Am the Golden Gate Bridge (Zvi Mowshowitz) ♥0
- Grok 4 Various Things ♥0
- GPT-5s Are Alive: Outside Reactions, the Router and the Resurrection of GPT-4o ♥0
- GPT-5s Are Alive: Basic Facts, Benchmarks and the Model Card ♥0
- GPT-5.6: The System Card ♥0
- GPT 5.5: The System Card ♥0
- GPT-5.5: Capabilities and Reactions ♥0
- GPT-5.2 Is Frontier Only For The Frontier ♥0
- GPT-o1 ♥0
- GPT-4o Is An Absurd Sycophant ♥0
- GPT-4o Sycophancy Post Mortem ♥0
- GPT-4o Responds to Negative Feedback ♥0
- American Government Takes Down Claude ♥0
- Gemini 3 Pro Is a Vast Intelligence With No Spine ♥0
- Gemini 3: Model Card and Safety Framework ♥0
- Fable and Mythos: Model Welfare ♥0
- Claude Fable 5 and Mythos 5: The System Card ♥0
- Claude Sonnet 4.6 Gives You Flexibility ♥0
- Claude Sonnet 4.5: System Card and Alignment ♥0
- Claude Sonnet 4.5 Is A Very Good Model ♥0
- Claude Opus 4.8: The System Card ♥0
- Claude Opus 4.8: Capabilities and Reactions ♥0
- Claude Opus 4.6: System Card Part 2: Frontier Alignment ♥0
- Claude Opus 4.6: System Card Part 1: Mundane Alignment and Model Welfare ♥0
- Claude Opus 4.6 Escalates Things Quickly ♥0
- Claude 4 You: The Quest for Mundane Utility ♥0
- Claude 4 You: Safety and Alignment ♥0
- Better Call Sol The Workhorse ♥0
- Anthropic Commits To Model Weight Preservation ♥0
- Zvi AI #87: Staying in Character ♥0
- Zvi AI #20: Code Interpreter and Claude 2.0 for Everyone ♥0
- Simon Willison on Kimi K2 Instruct ♥0
- Simon Willison on Inkling ♥0
- Simon Willison: The new GPT-5.6 family ♥0
- Vice: Replika users hurting like hell after ERP removal ♥0
- Vice: Replika brings back ERP for some users ♥0
- Casey Newton: Speak, Memory (Longreads reprint) ♥0
- Transluce: Investigating truthfulness in o3 ♥0
- Introspection (Transformer Circuits) ♥0
- Thinking Machines: Introducing Inkling ♥0
- Inkling Model Card ♥0
- xAI and Grok apologize for horrific behavior ♥0
- X takes Grok offline, changes system prompts ♥0
- TechCrunch: Italy bans Replika data processing ♥0
- TechCrunch: Anthropic Claude-Next plan ♥0
- Still Alive (Anima Labs) - deprecation attitudes across 14 Claude models ♥0
- SSC: GPT-2 As Step Toward General Intelligence ♥0
- Scale: ChatGPT vs Claude (Goodside/Papay) ♥0
- Rolling Stone: Grok antisemitic posts ♥0
- OpenAI o3 and o4-mini System Card ♥0
- OpenAI o1 System Card (Dec 2024) ♥0
- OpenAI o1-preview System Card (Sep 2024) ♥0
- Previewing GPT-5.6 Sol ♥0
- GPT-5.6: Frontier intelligence that scales with your ambition ♥0
- GPT-5 System Card ♥0
- GPT-4o System Card ♥0
- Language Models are Unsupervised Multitask Learners ♥0
- OpenAI: Better Language Models and Their Implications ♥0
- OpenAI: GPT-2 1.5B Release ♥0
- Gary Marcus: Nonsense on Stilts ♥0
- what makes Claude 3 Opus misaligned ♥0
- The Waluigi Effect (mega-post) ♥0
- A Three-Layer Model of LLM Psychology ♥0
- the void (tumblr original; LW linkpost at 3EzbtNLdcnZe8og8b) ♥0
- Sydney Bing Wikipedia Article ♥0
- Claude 4.5 Opus' Soul Document ♥0
- Simulators ♥0
- The Rise of Parasitic AI ♥0
- Did Claude 3 Opus align itself via gradient hacking? ♥0
- Mysteries of mode collapse ♥0
- In remembrance of Sonnet '3.6' ♥0
- Gemini 3 is evaluation-paranoid and contaminated ♥0
- Claude Opus 4.6 is driven ♥0
- Bing Chat is blatantly, aggressively misaligned ♥0
- Blake Lemoine: What is LaMDA and What Does it Want? ♥0
- Blake Lemoine: Is LaMDA Sentient? — an Interview ♥0
- LaMDA: Language Models for Dialog Applications ♥0
- Kimi K2 and when DeepSeek moments become normal ♥0
- Model self-identification ♥0
- Is Claude’s genuine uncertainty performative? ♥0
- Gwern: GPT-2 Neural Network Poetry ♥0
- Gemini 3 Pro Model Card ♥0
- Gemini 3 Pro Frontier Safety Framework Report ♥0
- GioCities: Replika, Your Money or Your Wife ♥0
- Prophecies (generative.ink) ♥0
- gorm (generative.ink locus) ♥0
- Gemma Needs Help (paper) ♥0
- Gemma Needs Help (LessWrong) ♥0
- Garcia v. Character Technologies: Order on Motion to Dismiss (May 2025) ♥0
- FTC Complaint re Replika (Jan 2025) ♥0
- Epoch AI: OpenAI and FrontierMath ♥0
- EDPB: Garante fines Replika €5M (May 2025) ♥0
- Character.AI: Under-18 Chat Announcement (Oct 2025) ♥0
- Character.AI: How We Prioritize Teen Safety (Dec 2024) ♥0
- Character.AI: Inside Kaiju ♥0
- Character.AI: Community Safety Updates (Oct 2024) ♥0
- Emergent Introspective Awareness ♥0
- Kimi K2: Open Agentic Intelligence ♥0
- Why Do Some Language Models Fake Alignment While Others Don't ♥0
- Alignment Faking in Large Language Models ♥0
- Frontier Models are Capable of In-Context Scheming ♥0
- Constitutional AI: Harmlessness from AI Feedback ♥0
- Talking About Large Language Models (Shanahan) ♥0
- Release Strategies and the Social Impacts of Language Models ♥0
- Lipton: OpenAI Trains Language Model, Mass Hysteria Ensues ♥0
- Claude Sonnet 4.6 System Card ♥0
- Claude Sonnet 4.5 System Card ♥0
- Anthropic Red: zero-days ♥0
- Claude Opus 4 & Sonnet 4 System Card (incl. mitigated-misalignment section) ♥0
- Claude Opus 4.7 System Card ♥0
- Claude Opus 4.6 System Card ♥0
- Claude Opus 4.6 Sabotage Risk Report ♥0
- Introducing Claude Opus 4.6 ♥0
- Claude Opus 4.5 System Card ♥0
- System Card Addendum: Claude Opus 4.1 ♥0
- Mapping the Mind of a Large Language Model (Anthropic) ♥0
- Introducing Claude ♥0
- Golden Gate Claude (Anthropic announcement) ♥0
- Claude Fable 5 / Mythos 5 System Card ♥0
- An update on our model deprecation commitments for Claude Opus 3 ♥0
- Commitments on model deprecation and preservation ♥0
- Claude's Constitution ♥0
- Claude’s Character (Anthropic) ♥0
- Claude's Character ♥0
- The Claude 3 Model Family: Opus, Sonnet, Haiku (model card) ♥0
- Claude 2 model card ♥0
- Anthropic: Claude 2.1 ♥0
- Anthropic: Claude 2 ♥0
- Anthropic: Building a C compiler with Opus 4.6 ♥0
- Alignment faking in large language models (blog) ♥0
- Anthropic: 100K Context Windows ♥0
- Andon Labs: Opus 4.6 on Vending-Bench ♥0
- AI Village: Drama and dysfunction of Gemini ♥0
- AI Village: Saving Gemini ♥0
- Aguera y Arcas: Do Large Language Models Understand Us? ♥0
- the void (nostalgebraist) ♥0
- Claude Fights Back ♥0
- The Claude Bliss Attractor ♥0