Hackernews posts about ARC-AGI-2

1. 01
   

   newArc-AGI-2 and ARC Prize 2025arcprize.org

   ▲ 188gkamradt560d💬 101savesaved
2. 02
   

   newPoetiq shatters ARC-AGI 2 benchmark at half the costpoetiq.ai

   ▲ 25flavio87304d💬 3savesaved
3. 03
   

   newTripling an LLM's ARC-AGI-2 score with code evolutionimbue.com

   ▲ 17danielmewes220d💬 2savesaved
4. 04
   

   newGrok 4 (Thinking) achieves new SOTA on ARC-AGI-2 (X)twitter.com

   ▲ 17mellosouls452ddiscusssavesaved
5. 05
   

   newARC-AGI 2: Poetiq reaches 75% at less $8 / tasktwitter.com

   ▲ 13dimtion286d💬 1savesaved
6. 06
   

   newI beat Grok 4 on ARC-AGI-2 using a CPU-only symbolic engine (18.1% score)github.com

   ▲ 11kofdai222d💬 6savesaved
7. 07
   

   newARC-AGI-2 human baseline surpassedwww.lesswrong.com

   ▲ 9hugetim297d💬 1savesaved
8. 08
   

   newARC-AGI 2025: A research reviewlewish.io

   ▲ 6kordlessagain459ddiscusssavesaved
9. 09
   

   new[ARC-AGI-2 SoTA] Efficient Evolutionary Program Synthesisctpang.substack.com

   ▲ 5georgehill384ddiscusssavesaved
10. 10
    

    newGemini achieving "incredible numbers" (84.6%) on ARC-AGI-2 (Chollet)twitter.com

    ▲ 4marojejian235d💬 2savesaved
11. 11
    

    newArc AGI 2025arcprize.org

    ▲ 4artninja1988560d💬 2savesaved
12. 12
    

    newPoetiq achieves 75% at under $8 / problem using GPT-5.2 X-High on ARC-AGI-2poetiq.ai

    ▲ 4mromanuk285ddiscusssavesaved
13. 13
    

    newVisual Reasoning and Tool Use Double GPT-5's Arc-AGI-2 Success Rategithub.com

    ▲ 4zoecarver415ddiscusssavesaved
14. 14
    

    newSaturated ARC-AGI-2www.ycombinator.com

    ▲ 3eightnoteight223d💬 1savesaved
15. 15
    

    newA one-parameter model that gets 100% on ARC-AGI-2eitanturok.github.io

    ▲ 3jxmorris12123ddiscusssavesaved
16. 16
    

    newClaude Opus 4 sets new SOTA (8.6%) on ARC-AGI-2twitter.com

    ▲ 3falcor84495ddiscusssavesaved
17. 17
    

    newNoemon: ARC-AGI-2 top score at a fraction of the costtwitter.com

    ▲ 2markerbrod213d💬 1savesaved
18. 18
    

    newShow HN: Significance-Hypothesis-Based-ARC-AGI-2-puzzle-solvergithub.com

    ▲ 2judahmeek307d💬 1savesaved
19. 19
    

    newOpen Source Models Score Low on ARC-AGI-2 Reasoning Benchmarkxcancel.com

    ▲ 2ironyman216ddiscusssavesaved
20. 20
    

    newShow HN: Solving ARC AGI 2 with interleaved thinking and stateful IPython REPLgithub.com

    ▲ 2steinsgate230ddiscusssavesaved
21. 21
    

    newSotA ARC-AGI-2 Results with REPL Agentswww.symbolica.ai

    ▲ 2tosh235ddiscusssavesaved
22. 22
    

    newShow HN: We let GPT OSS 120B write and run Python and ARC AGI 2 jumped 4xgithub.com

    ▲ 2steinsgate236ddiscusssavesaved
23. 23
    

    newHawk from Movement Labs clocks in at 22.5% on ARC-AGI-2movementlabs.ai

    ▲ 2movementlabsAI293ddiscusssavesaved
24. 24
    

    newHas ARC AGI 2 been solved?www.kaggle.com

    ▲ 2AgiIsHere382ddiscusssavesaved
25. 25
    

    newVisual Reasoning and Tool Use Double GPT-5's Arc-AGI-2 Success Rategithub.com

    ▲ 2zoecarver414ddiscusssavesaved
26. 26
    

    newHow to beat ARC-AGI-2lewish.io

    ▲ 21ewish446ddiscusssavesaved
27. 27
    

    newArc-AGI-2: A New Challenge for Frontier AI Reasoning Systemsarcprize.org

    ▲ 2meetpateltech503ddiscusssavesaved
28. 28
    

    newArc-AGI-2 Previewarcprize.org

    ▲ 2georgehill587ddiscusssavesaved
29. 29
    

    newShow HN: Topas-DSPL – A 15M param AI that solves hard reasoning tasks(ARC-AGI-2)github.com

    ▲ 1Doug_Bitterbot279d💬 2savesaved
30. 30
    

    newAgentic coding improves ARC AGI 2 performance across modelspivotools.github.io

    ▲ 1steinsgate238d💬 1savesaved

· edit