Deniz Bayazit
@bayazitdeniz.bsky.social
📤 8
📥 13
📝 7
#NLProc
PhD student @EPFL
#interpretability
reposted by
Deniz Bayazit
Badr AlKhamissi
18 days ago
🚀 Excited to share a major update to our “Mixture of Cognitive Reasoners” (MiCRo) paper! We ask: What benefits can we unlock by designing language models whose inner structure mirrors the brain’s functional specialization? More below 🧠👇
cognitive-reasoners.epfl.ch
2
29
10
1/🚨 New preprint How do
#LLMs
’ inner features change as they train? Using
#crosscoders
+ a new causal metric, we map when features appear, strengthen, or fade across checkpoints—opening a new lens on training dynamics beyond loss curves & benchmarks.
#interpretability
about 1 month ago
2
14
6
you reached the end!!
feeds!
log in