ACCORD: Action-Conditioned Contextual Grounding for Language Agents
Researchers have introduced ACCORD, a new framework designed to improve the performance of language agents by enabling them to better ground their actions in observed environmental context. ACCORD addresses the issue of underspecified instructions by actively probing for missing information and integrating relevant context from the agent's history before each action. This approach significantly enhances task completion rates, showing improvements of up to 20.6 points with GPT-5-mini on the AppWorld benchmark, and also demonstrates gains with other models like Claude-4.5-sonnet and Qwen3.5-27B-FP8. AI
IMPACT Enhances LLM agent performance by improving context grounding and task completion.