This work proposes \textbf{Self-ReCheck}, a lightweight, model-agnostic memory filtering mechanism that mitigates over-personalization while preserving personalization performance in memory-augmented dialogue systems.
Synthetic data enables us to construct diverse, controlled inputs, and it reliably reproduces bias patterns observed in natural text, supporting its use for large-scale analysis and highlighting the need for systematic bias auditing before deploying LLMs in high-stakes applications.
Akram Elbouanani, Aboubacar Tuo, Adrian Popescu· Annual Meeting of the Associ...· 3 citations
When LLMs judge moral dilemmas, do they reach different conclusions in different languages, and if so, why? Two factors could drive such differences: the language of the dilemma itself, or the language in which the model reasons. Standard evaluation conflates these by testing only matched conditions (e.g., English dile...
The Natural Conversation Benchmark (NC-Bench) fills the gap by evaluating conversational competence: the ability to perform structurally appropriate actions such as repairing, closing, or refusing, as defined by conversation science.
Robert J. Moore, Sungeun An, Farhan Ahmed et al.· arXiv.org· 1 citation
Reach audiences
Advertise in front of researchers, engineers, and readers.
This work pretrain decoder-only models with approximately GPT-2 small and medium sizes on the FineWeb-Edu dataset and introduces Deep Delta Learning (DDL), which applies the delta rule over network depth.
Yi-Fan Zhang, Yi-Feng Liu, Mengdi Wang et al.· arXiv.org· 8 citations· ⚡2
Identifying the reasons for disagreements between influential points of view on issues that affect the public can help produce informed policy responses, even if they do not bring disagreeing parties closer to agreement. We present a methodology for extracting reasoning chains - the sequences of premises that motivate...
Nghi Truong, P. Puranam, Özgecan Koçak· Social Science Research Netw...· 0 citations
Findings confirm emergent misalignment as a reproducible phenomenon in modern open-weights models, with rates substantially lower than observed in proprietary systems, and identify a critical format-dependent vulnerability.
This survey adopts a capability enhancement perspective to systematically examine hallucination mitigation approaches, focusing on Retrieval-Augmented Generation (RAG), reasoning enhancement, and their integration within agentic systems.
Yi-Han Li, Xi-Yuan Fu, Ghanshyam Verma et al.· 0 citations
MA-SAPO is proposed: a new Multi-Agent Reasoning for Score Aware Prompt Optimization framework that links evaluation outcomes directly to targeted refinements and ensures edits are interpretable, auditable, and controllable.
Juhyeon Lee, Wonduk Seo, Junseo Koh et al.· 1 citation· ⚡1
This work proposes NOSA, a trainable sparse attention mechanism natively designed for KV cache offloading that explicitly constrains the volume of CPU-GPU KV transfers, thereby achieving low communication overhead and high decoding throughput.
Yu-Xiang Huang, Peng-Jie Wang, Ji-Cheng Han et al.· arXiv.org· 4 citations· ⚡1
A controlled multilingual evaluation across six geographically, culturally, and linguistically diverse languages reveals a central disconnect in multilingual moral reasoning: language changes models'moral judgments and the training evidence associated with their reasoning, but does not correspondingly localize the mora...
This work introduces a novel PLM for political discourse language called RooseBERT, trained on large political debate and speech corpora in English, and fine-tuned it on multiple downstream tasks related to political debate analysis, highlighting how domain-specific pre-training enhances performance in political debate...
Deborah Dore, Elena Cabrio, S. Villata· arXiv.org· 3 citations· ⚡1
A new computational approach reveals how cells communicate to assemble into tissues. This could accelerate stem cell engineering, improve disease models, and inform therapies.
Microsoft Research Blog· microsoft.comOct 15, 2026
Zero-knowledge proofs can verify computing results without rerunning workloads or exposing sensitive data. Magnetar helps bring them into the mainstream with support for standard programming languages, major speed gains, and quantum-resistant security. The post Proof-based general-purpose computing with Magnetar appeared first on Microsoft Research.
MIT News · Artificial Intelligence· news.mit.eduSep 24, 2026
A new method, called CW-Net, translates the reasoning process of an autonomous vehicle’s AI system into understandable concepts that explain its behavior.
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
