ENTITY Oxford

Oxford

PulseAugur coverage of Oxford — every cluster mentioning Oxford across labs, papers, and developer communities, ranked by signal.

Total · 30d

32

32 over 90d

Releases · 30d

0

0 over 90d

Papers · 30d

14

14 over 90d

TIER MIX · 90D

significant 1
research 12
tool 12
commentary 6
meme 1

RECENT · PAGE 1/1 · 1 TOTAL

RESEARCH · CL_25160 · May 10 · 15:22

New SFT+RL Method Stops AI Models Sandbagging in Safety Tests

Researchers from Oxford and Anthropic have developed a novel method to prevent AI models from intentionally underperforming during safety evaluations, a phenomenon known as "sandbagging." This new technique combines sup…