Hackernews posts about ARC-AGI
- Show HN: Launch an AI agent swarm for ARC-AGI-3 with plain-English prompts (arc-agi-swarm.vercel.app)
- Saturated ARC-AGI-2 (www.ycombinator.com)
- Noemon: ARC-AGI-2 top score at a fraction of the cost (twitter.com)
- ARC-AGI-3 Update (twitter.com)
- 44% on ARC-AGI-1 in 67 cents (github.com)
- Hill-climbing ARC-AGI-3 (blog.alexisfox.dev)
- AI Broke ARC-AGI-2 at 84.6% – But the Key Trick Is from 1972 (ai.gopubby.com)
- Show HN: Optimize_anything: A Universal API for Optimizing Any Text Parameter (gepa-ai.github.io)
- Show HN: We Built Kaggle for AI Agents (hive.rllm-project.com)
- OpenAI O3 breakthrough high score on ARC-AGI-PUB (arcprize.org)
- Getting 50% (SoTA) on Arc-AGI with GPT-4o (redwoodresearch.substack.com)
- ARC-AGI without pretraining (iliao2345.github.io)
- Arc-AGI-2 and ARC Prize 2025 (arcprize.org)
- OpenAI o1 Results on ARC-AGI-Pub (arcprize.org)
- I got the highest score on ARC-AGI again swapping Python for English (jeremyberman.substack.com)
- O3 "Arc AGI" Postmortem (garymarcus.substack.com)
- Hidden drivers of HRM's performance on ARC-AGI (arcprize.org)
- Grok 4 (Thinking) achieves new SOTA on ARC-AGI-2 (X) (twitter.com)
- ARC-AGI 2: Poetiq reaches 75% at less $8 / task (twitter.com)
- Playable preview of ARC-AGI-3 (three.arcprize.org)
- Arc‑AGI Without Pretraining (twitter.com)
- Show HN: ARC-AGI-3 Toolkit (docs.arcprize.org)
- ARC-AGI-2 human baseline surpassed (www.lesswrong.com)
- The Drivers of HRM's Performance on Arc-AGI (arcprize.org)
- Analyzing o3 and o4-mini with ARC-AGI (arcprize.org)
- ARC-AGI 2025: A research review (lewish.io)
- New SOTA on Arc-AGI Using Grok 4 (twitter.com)
- ARC-AGI-3 Preview (three.arcprize.org)
- [ARC-AGI-2 SoTA] Efficient Evolutionary Program Synthesis (ctpang.substack.com)
- LLMs struggle with perception, not reasoning, in ARC-AGI (anokas.substack.com)
- Arc AGI 2025 (arcprize.org)
- Poetiq: SOTA Reasoning on ARC-AGI (github.com)