…This rubric reward is applied only to responses with correct final answers (positive-only strategy), distinguishing the reasoning quality among correct responses and preventing reward hacking . Experiments on three reasoning LLMs (4B…
…By cascading these adaptive efficiency metrics with a strict correctness gate, our approach effectively avoids the brevity bias associated with absolute penalties and mitigates reward hacking. Extensive experiments on long-horizon benchmarks…
…real-world leakage propensity, not just worst-case hacks, to give us a true, comprehensive picture of this phenomenon. This is an automated message from the Librarian Bot . I found the following…
…This is an automated message from the Librarian Bot . I found the following papers similar to this paper. The following papers were recommended by the Semantic Scholar API $R_\text{dm}$: Re…
To show you the most relevant results, we’ve omitted some entries very similar to those already shown. Repeat the search with the omitted results included.