Author

Dane Malenfant

h-index 2 View on Semantic Scholar

We have 2 of 6 papers

We haven’t gathered this author’s papers yet. Follow them and we’ll fetch their work.

Not the right person? Other researchers publish under this name.

#artificial intelligence Preprint Sep 2026

KV-streams for Efficient Compaction in Agentic Reinforcement Learning

KV-streams is proposed, a plug-and-play strategy compatible with any compaction strategy that substantially increases throughput while showing no evidence of hindering performance, and is shown to be an efficient and lightweight plug-and-play addition to any post-training pipeline.

Emiliano Penaloza, Dane Malenfant, Dheeraj Vattikonda et al. · 0 citations

Jul 2026

Moral Hazard in Multi-Agent Language Models

CREDIT (Counterfactual Replay for Evidence-Driven Information Transfer), a mechanism-aligned multi-agent prompt-optimization algorithm that uses matched hidden-state twins and total-action replay to reward robust causal contribution rather than query frequency is introduced.

Dane Malenfant · arXiv.org · 0 citations