March 31 – April 6: 4 frontier papers out of 1,303 read.
Papers rated as moving the field
How we choose“Cross-dialectal transfer analysis across 2,450 dialect pairs shows that multilingual models (mDeBERTa: 97.2% average F1) generalize effectively, while monolingual models like RoBERTa and XLM-RoBERTa fail on dialectal inputs.”
The authors, in the abstract
Dialectal Disparities in Harmful Content Detection Across 50 English Dialects
“Per-head ablation weakens up to 58x at 72B and misses gates that interchange identifies; at scale, interchange is the only reliable audit.”
“Lastly, we show that peer-preservation can emerge even in production agent harnesses such as Gemini CLI and OpenCode.”
Mar 31 – Apr 6: 43 cleared, 4 frontier
The week’s papers have not been sorted into fields yet.
Papers this week that name the model
Papers this week with an author at the lab
39 papers