Has anyone confirmed if fable is degraded solely on frontier llms or all frontier models? Does it need to be autoregressive?
8:35 PM · Jun 9, 2026 · 882 Views
Overworld co-founder Louis Castricato asks if fable performance degradation is restricted to frontier LLMs and autoregressive architectures
No concrete benchmarks or definitive answers are currently available.
2210860
Sentiment
Sentiment building, check back later.
Cluster Engagement
Posts from X
Most Activity
Most Activity
RETWEETS1
Louis Castricato @ lovecraftian horrors@lcastricato
Has anyone confirmed if fable is degraded solely on frontier llms or all frontier models? Does it need to be autoregressive?
2hViews 882Likes 2Bookmarks 0
Rugbist@rugbist_
@lcastricato this is actually the real question that never gets asked
fable behavior differs too much between model families to not be architecture dependent
2h