CogGym Finds Bigger, Newer Models Behave More Like Humans — But Even More Like Each Other
teortaxesTex · x · 2026-10-03
TeortaxesTex highlights CogGym, a framework for evaluating the similarity of cognitive behavior between AI and humans. Two key takeaways: larger, more recently released models show higher behavioral alignment with humans; yet models are more similar to each other than to us. The poster calls it a win for natural abstractions research while flagging the counterintuitive finding.
More from Research
- Schmidhuber Replies to the Pope With His 2008 Curiosity Paper on Compression Progress — prajdabre · 2026-10-03
- CMU team releases SMDD-Bench: 502 small-molecule drug design tasks for RL agents — willcb · 2026-10-03
- MBZUAI's Ego2Act Benchmark Shows Video Models Skip Steps and Break Physical Plausibility — MBZUAI · 2026-10-03
- First NECB2026 draws 360+ researchers, 190+ abstracts over two days — KevinKaichuang · 2026-10-03
- TLAPS-Bench debuts: proving real-world TLA+ specs remains hard for frontier AI — tianyin_xu · 2026-10-03
- Looped LMs could use 3x fewer parameters and a 3x smaller KV cache, matching standard Transformers — ChengleiSi · 2026-10-03