phenomenon · first seen 7 Oct 2026
knowledge conflict
Also: context-memory conflict, knowledge conflict in language models, contextual knowledge reliance in LLM translation
Situations where a model's parametric knowledge or retrieved sources contradict each other.
Mentions per week
PapersCodeDiscussion
Earliest sources
- The Attribution Blind Spot: Layerwise Trajectory Diagnostics for Source Reliance in Retrieval-Augmented Language ModelsPapers · 7 Oct 2026 · Zhe Yu, Wenpeng Xing, Yunzhao Wei, Bo Yang, Chen Ye, Gaolei Li, Meng Han
- Accurate but Not Humble: Evaluating Epistemic Humility in LLM Agents under Knowledge ConflictPapers · 8 Oct 2026 · Kaiser Sun, Bernal Jimenez Gutierrez, Hongjun Liu, Jingyu Zhang, Jie Gao, Mark D…
- Which Knowledge Should LLMs Trust? Uncovering a Generalizable Steering Direction Across Knowledge ConflictsPapers · 14 Oct 2026 · Yibin Lou, Yi Hu, Jianhui Chen, Chenpeng Wang, Muhan Zhang, Liangming Pan
- Accept or Repair: Preserving Passage-Supported Answers in Contrastive DecodingPapers · 21 Oct 2026 · Jungseob Lee, Seungyoon Lee, Hyeonseok Moon, Seongtae Hong, Suhyune Son, Sugyeon…
- Do We Need an LLM to Know Why Sources Conflict? Decision Models for Knowledge-Conflict Typing in RAGPapers · 27 Oct 2026 · Parv Khatri, Harshul Jain, Mehak Kapoor, Tanya Sah
- Are These Programmers Ripe Enough to Eat? Evaluating How LLMs Handle Lexical Knowledge Conflicts in TranslationPapers · 6 Nov 2026 · Dongxu Wu, Sara Court, Micha Elsner
Latest sources
- Are These Programmers Ripe Enough to Eat? Evaluating How LLMs Handle Lexical Knowledge Conflicts in TranslationPapers · 6 Nov 2026 · Dongxu Wu, Sara Court, Micha Elsner
- Do We Need an LLM to Know Why Sources Conflict? Decision Models for Knowledge-Conflict Typing in RAGPapers · 27 Oct 2026 · Parv Khatri, Harshul Jain, Mehak Kapoor, Tanya Sah
- Accept or Repair: Preserving Passage-Supported Answers in Contrastive DecodingPapers · 21 Oct 2026 · Jungseob Lee, Seungyoon Lee, Hyeonseok Moon, Seongtae Hong, Suhyune Son, Sugyeon…
- Which Knowledge Should LLMs Trust? Uncovering a Generalizable Steering Direction Across Knowledge ConflictsPapers · 14 Oct 2026 · Yibin Lou, Yi Hu, Jianhui Chen, Chenpeng Wang, Muhan Zhang, Liangming Pan
- Accurate but Not Humble: Evaluating Epistemic Humility in LLM Agents under Knowledge ConflictPapers · 8 Oct 2026 · Kaiser Sun, Bernal Jimenez Gutierrez, Hongjun Liu, Jingyu Zhang, Jie Gao, Mark D…
- The Attribution Blind Spot: Layerwise Trajectory Diagnostics for Source Reliance in Retrieval-Augmented Language ModelsPapers · 7 Oct 2026 · Zhe Yu, Wenpeng Xing, Yunzhao Wei, Bo Yang, Chen Ye, Gaolei Li, Meng Han