Tool-using language-model agents often execute actions that their domain policies forbid. We present LeanGuard, a runtime that checks every proposed tool call against a policy written in Lean4 before the call reaches the environment. Policies combine state predicates with scoped past-time temporal operators over the ex...
Sreehari Sankar, Debargha Ganguly, Vikash Singh et al.· 0 citations
RLVR trains language models to reason by rewarding correct final answers. A widely used algorithm, group-relative policy optimization (GRPO), gives every token of a solution the same credit, so dense process rewards have been proposed to credit intermediate steps, for example with a Monte-Carlo reward that scores each...
Vikash Singh, Wei-Cong Chen, Debargha Ganguly et al.· 0 citations
Post-training raises the 4B model's per-attempt accuracy and four-sample all-correct reliability and reduces average SMT calls from 4.10 to 2.52 and tokens per attempt from 18.97k to 12.72k.
Debargha Ganguly, Wei-Cong Chen, Mohsen Hariri et al.· 4 citations
Applied to six watermarking schemes, three open-weight generators, eleven languages spanning four scripts and eight typological families, and both base and instruction-tuned regimes, the framework reveals failure modes that single-language single-paradigm evaluation cannot surface.
Alexander Nemecek, Osama Zafar, Debargha Ganguly et al.· 1 citation
These findings show that safe-looking behavior alone is insufficient; live scam resistance must separately measure intervention, timing, structural localization, and false-positive behavior.
Yu-Qiao Xu, Osama Zafar, Alexander Nemecek et al.· 0 citations
The Infinite Impostor is introduced, an attack model in which an autonomous agent interposes itself between two parties who already trust each other, hijacking an existing relationship rather than building a new one from scratch, and the governance tensions that arise when platforms become the regulatory substrate of d...
Osama Zafar, Alexander Nemecek, Erman Ayday· arXiv.org· 0 citations
It is argued that neither the objections nor the assurances can currently be verified and that this unverifiability, rather than watermarking itself, is the substantive governance failure.
Alexander Nemecek, Vipin Chaudhary, Erman Ayday· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.