The paper introduces PACE, a dataset that tests whether personalized assistants can detect hidden conflicts in user requests by integrating implicit, egocentric knowledge from a knowledge base. PACE pairs user requests with persona-specific facts, requiring models to retrieve contextual evidence before deciding if a request is inappropriate. The authors also propose PaceMaker, a multi-agent system that improves conflict detection by coordinating query reformulation, multi-hop graph traversal, and conflict-aware filtering, outperforming existing methods on the PACE benchmark.
arXiv:2606. 20245v1 Announce Type: new Abstract: Large language models (LLMs) have achieved strong performance across a wide range of language-based tasks by leveraging both extensive parametric knowledge and in-context learning ability, enabling them to incorporate external information provided in the input prompt.
KC-Bench is a dynamic interactive benchmark designed to evaluate how large language model agents reconcile user instructions, internal knowledge, and real‑time environmental observations. It contains 238 manually curated multi‑turn tasks that test world‑knowledge conflicts, input inconsistencies, and multi‑source temporal conflicts, using a user simulator, stateful tools, deterministic environment assertions, an open‑source natural‑language evaluator, and human trajectory verification. Evaluation of nine models—including DeepSeek‑V4‑Flash, GLM‑5.2, and MiniMax‑M3—reveals significant cross‑domain variation, with no model reliably handling factual correction, identity consistency checking, and temporal conflict resolution across all settings, and shows that missed conflicts can propagate to tool calls or synthetic protected‑data flows.
The paper introduces a taxonomy of six types of contextual knowledge conflicts—factual, inferential, temporal, granularity, perspective, and ambiguity—and presents the ContextConflict dataset with 5,781 samples covering reasoning and summarization tasks. Experiments on nine large language models reveal that current models struggle to resolve these conflicts, exhibit a bias toward earlier evidence, and show latent awareness of conflicts in their internal representations. The authors propose a training‑free, label‑free steering method that adjusts activations to better incorporate evidence, consistently improving reasoning accuracy and producing higher‑quality, balanced summaries on the dataset.
The paper examines how large language models resolve conflicts that arise within contextual knowledge, rather than between internal knowledge and external context. It introduces a taxonomy of six contextual conflict types and presents the ContextConflict dataset with 5,781 samples covering reasoning and summarization tasks. Experiments on nine LLMs reveal persistent shortcomings in conflict resolution, uncover a bias toward earlier evidence, and propose a training‑free steering method that improves accuracy and summary quality.
KC-Bench is a dynamic interactive benchmark designed to evaluate how large language model agents reconcile user instructions, internal knowledge, and real‑time environmental observations. It consists of 238 manually curated multi‑turn tasks that test world‑knowledge conflicts, input inconsistencies, and multi‑source temporal conflicts, using a user simulator, stateful tools, deterministic environment assertions, an open‑source natural‑language evaluator, and human trajectory verification. Evaluation of nine models—including DeepSeek‑V4‑Flash, GLM‑5.2, and MiniMax‑M3—reveals significant cross‑domain variation, with no model reliably handling factual correction, identity consistency, and temporal conflict resolution across all settings, and shows that missed conflicts can propagate to tool calls or synthetic protected‑data flows.
arXiv:2608. 13921v1 Announce Type: new Abstract: LLM agents increasingly maintain personal memory across sessions, but it can conflict.
LLM agents increasingly maintain personal memory across sessions, but it can conflict. Preferences depend on context, behavior evolves, and sources can conflict. When a query lacks context, time, or s...
The paper introduces UC-Bench, a human‑annotated benchmark for detecting user‑side implicit conflicts in Human‑LLM dialogue, a problem largely overlooked compared to LLM‑side conflicts. Experiments show current LLMs struggle with these conflicts, especially when they stem from implicit incompatibilities in dialogue history. To address this, the authors propose SynUC, a constraint‑guided data synthesis method that generates a new training set, UC‑Data, which improves performance of lightweight LLMs on UC‑Bench compared to larger general‑purpose models and existing synthesis approaches.
The paper introduces CONFLICTGUI, a benchmark that tests GUI agents on instruction-internal and instruction‑GUI context conflicts, revealing that many agents over‑comply with infeasible instructions. To address this, the authors propose CONFLICTGUARD, an inference‑time framework that couples a feasibility verification protocol with a conditional action modulation mechanism, enabling agents to assess instruction logic and GUI evidence before acting. Experiments on five agents show that CONFLICTGUARD significantly improves conflict‑task success while maintaining normal task performance.
arXiv:2605. 17301v2 Announce Type: replace-cross Abstract: Retrieval-Augmented Generation (RAG) systems implicitly assume mutual consistency among retrieved documents -- an assumption that frequently fails in practice.
The paper introduces TRACE, a fine‑tuning framework for Retrieval‑Augmented Generation (RAG) that addresses conflicts between retrieved knowledge and a model’s internal knowledge. TRACE uses multi‑agent debate traces to identify correct and incorrect candidates and answer‑shift patterns, providing fine‑grained supervision for reliable knowledge‑source selection. It also incorporates an answer‑completeness regularization mechanism to prevent empty, overly short, or prematurely terminated responses, thereby improving robustness against misleading retrieved content and enhancing answer quality.
The paper introduces a taxonomy of six user challenge types and a four-layer framework to analyze how large language models respond to user disagreement. Using a dataset of 2,310 challenge scenarios and 32,340 responses from 14 models, the study finds that models often validate users (85%) while still maintaining their original claim (65%). It also reports that models frequently apologize (33%) and transfer authority in advice contexts, with significant variation across model types and task domains.