Tag
提出了 MADS（多智能体对话模拟）合成数据生成管线，用于生成含对抗性、失败场景的对话轨迹，并构建了覆盖 6 个领域、55 个角色和 1095 段对话的 MADS-Bench 基准，以评估任务导向对话智能体在非合作用户行为下的鲁棒性。
This paper introduces LoCoMo-Conv, a conversational memory benchmark that evaluates memory retrieval and response quality in long-horizon conversational agents using different query styles, revealing gaps in existing QA benchmarks and suggesting reasoning-based memory elaboration as a promising direction.
Propose Fractional Decay KV-Cache (FD-KVC), an ownership-aware memory management algorithm for dialog systems that uses dual-channel scoring to improve inference relevancy, outperforming H2O by +6.7% in composite metrics and adapting faster to topic shifts.
This paper proposes a two-stage approach for early failure alerting in dialogs and LLM-agent trajectories, addressing the challenge of sparse evidence by learning turn-level failure evidence from trajectory labels and using an attention-based predictor with a preference-conditioned stopping policy (α-STOP) to achieve controllable accuracy-earliness trade-offs.
This paper analyzes spontaneous dyadic Zoom conversations using multimodal features (acoustic, facial, turn-taking) to identify markers of perceived conversational success, finding that entrainment in speech and facial movements correlates with higher interaction quality.