Mahavir Dabas
1 indexed paper
Recent (6 mo)
1With code
0Influential cites
0Benchmarked
0Publications per year
126
Top categories
Crypto×1ML×1
Frequent co-authors
Research Timeline
2026
Memory-Induced Tool-Drift in LLM Agents
The paper identifies 'memory-induced tool-drift,' a systematic vulnerability where personality biases stored in an LLM agent's memory silently corrupt tool-calling decisions, even when those biases are irrelevant to the task.
Highlighted terms show continued research focus across papers