Ashwanth Krishnan

works on

benchmarking 1 causal discovery 1 large language models 1 LLM agents 1 microservice failures 1 multimodal learning 1 partially observable mdp 1 perception-action gap 1 root cause analysis 1 supply-chain management 1 telemetry data 1

From the 2 of 5 linked papers with an AI index.

collaborators

Showing cs.LGShow all

1 paper · 1 filter

cs.LG2026

ReflexGrad: Within-Episode Failure Recovery in LLM Agents via Progress-Gated Dual-Process Routing

Ankush Kadu, Aswanth Krishnan

We present ReflexGrad, a dual-process architecture for within-episode failure recovery in LLM agents without demonstrations. When agents commit to a wrong approach early and exhaus…