Get Started
Home
Topics
Search
Library
273 papers · page 11 of 28

Research archive

Every paper covered by r*cap, newest first.
Verify Before You Distill: Prompt-Level Teacher Gating for On-Policy Distillation
Code Generation · Added Sep 9 · 6:50
Three Ways "Our Agent's Memory Still Works" Can Be Wrong
Agents · Added Sep 8 · 12:45
A Decodable Feature Is an Audit Lead, Not a Capability Contract
Evaluation · Added Sep 8 · 14:55
Three Independent Reasons a Robot Success Rate Can Mislead You
Evaluation · Added Sep 8 · 12:09
Your Model's Chain of Thought Is a Sensor, Not a Security Boundary
Alignment · Added Sep 8 · 14:56
A Hallucination Score Is Two Different Things
Evaluation · Added Sep 8 · 11:56
Two Different Failures Look Identical in Your Audio Eval
Audio/Speech · Added Sep 8 · 14:03
Two Cache Risks: Replay Failures and Hidden Grounding Loss
Evaluation · Added Sep 8 · 13:25
Before You Trust an LLM Personality Score
Evaluation · Added Sep 8 · 13:04
Agent Memory Controls Must Follow Consequences, Not Labels
Agents · Added Sep 8 · 10:35
12345678910111213141516171819202122232425262728