Can a model predict acceptance from the paper alone? | PaperScorer · 2026.10.22
Why do full-duplex voice models fall apart in long conversations? | FD-LHB · 2026.10.23
Why is AI-written text so hard to read? | Hard AI Text · 2026.10.21
Does a minimum count in the schema make LLMs pad their lists? | Padding to k · 2026.10.22
Rerun the same LLM judge. Does the verdict hold? | Behind Aggregate Scores · 2026.10.21
Can agents speed up real code and prove nothing changed? | Verified Optimization · 2026.10.20
Fix one caught collusion: does the uncaught kind shrink too? | Corrective Transfer · 2026.10.22
Show LLM agents what others picked. Do they herd more than people? | LLM Herding · 2026.10.22
Did RL teach reasoning models to say Wait? | Inherited, Not Invented · 2026.10.22
Can a human–AI loop fill every degree of Lehmers problem? | Generalized Boyd · 2026.10.21
Can two tools make a coding agents LEGO buildable? | BrickWright · 2026.10.22
Does standard softmax hide a bug? | Softmax Fix · 2026.10.22
Can you pick the data schedule without the learning rate? | Data × LR Schedules · 2026.10.21
Where do LLMs learn to recommend AI? | Pro-AI Prominence · 2026.10.21
