The paper introduces Neighborhood Elites (NE), a quality‑diversity search method for generating answer‑preserving rewrites in question answering. NE explicitly controls two dimensions—type of edit and explicitness of support—to create diverse perturbations that improve uncertainty quantification. Across five extractive‑QA datasets and five instruction‑tuned LLMs, NE outperforms other constructors, achieving higher AUROC and AUPRC scores and surpassing semantic entropy baselines.
By Mesut Ceylan, Arda Baris Basaran, Elgar Fleisch, Filipe Barata
The paper introduces Prompt-agnostic Language Models, focusing on train‑time techniques to embed prompt robustness during instruction tuning. It shows that mixing paraphrases in training causes significant gradient interference—57% of per‑paraphrasing gradients conflict—leading to the development of two gradient‑merging methods, Prompt‑TIES and Prompt‑DARE, which resolve this interference. Experiments across four models demonstrate that these methods substantially reduce per‑example variance and improve local robustness with minimal impact on overall accuracy.
By Frederic Sadrieh
The paper shows that classical consensus theory, which relies on network topology and convex update rules, fails for large language models (LLMs). By measuring token‑level updates across eight instruction‑tuned models and 73,080 exchanges, the authors find that LLMs violate convexity, exhibit negative weights, and are driven by internal update bias rather than network structure. Simulations with 10,000 agents confirm that algebraic connectivity explains almost none of the outcome variance, leading populations to converge on confidently false claims regardless of their network topology.
By Sahil Rajesh Dhayalkar
arXiv:2610. 37828v1 Announce Type: cross Abstract: An AI developer often has several candidate versions of a system and can deploy a mixture of them, sending each request to a randomly chosen version, as in A/B testing.
By Nicholas Teh, Paul W. Goldberg
Vendster is an evaluation arena that studies how the bias and prediction‑interval width of time‑series forecasters influence the outcomes of buyer‑vendor discount negotiations conducted by large language model (LLM) agents. In its reference instance, Vendster‑M5, forecasts generated by LLMs increased sales target attainment by up to 17 percentage points, yet traditional forecast accuracy metrics did not predict negotiation success. Controlled experiments revealed that under‑prediction bias significantly improves outcomes for agents that react strongly to forecasts, while interval width has no notable effect when bias is zero.
By Charin Polpanumas
The paper addresses the conflict between generalization and continuous adaptation in few‑shot and continual learning. It proposes a Complementary Learning System (CLS) that separates a fast Short‑Term Memory (STM) for memorizing recent inputs from a slow Long‑Term Memory (LTM) that provides robust, generalized perceptions. Experiments on image classification show the CLS model resists forgetting in domain‑incremental continual learning and outperforms other continual learners in few‑shot settings, though overall accuracy is slightly lower than some existing methods.
By David Rawlinson, Gideon Kowadlo
SCARL (Sequence Confidence-Aligned Reinforcement Learning) addresses a misalignment between a reinforcement learning policy’s confidence and its reward signal in language-model reasoning. By reshaping each advantage with an alignment gap that reflects the difference between a response’s group-relative confidence and the outcome it certifies, SCARL drives responses with equal advantage toward equal confidence and mitigates entropy collapse. Experiments on Qwen3‑4B‑Base show a 2.4‑point lift in Pass@32 over DAPO and consistent improvements across multiple estimators and backbones while preserving policy entropy.
By Zhenghong Ke, Jingbo Zhou, Hua Wu
The paper introduces a method for measuring political positions in data‑sparse regions by treating a large language model as a single fallible rater within a panel, analogous to expert surveys. It demonstrates that this panel approach yields reliable agreement across multiple models and reveals that disagreements often stem from differing interpretations of the regime rather than factual errors. The authors provide a full rubric, prompt, and data set, using the Middle East and North Africa as a worked example.
By Tarek Gara
The paper investigates whether federated learning harms the faithfulness of Grad‑CAM explanations in deepfake detection. Using an EfficientNet‑B4 model on FaceForensics++ data distributed across five clients, the study finds that explanation quality is non‑monotonic: near‑IID and moderate heterogeneity yield explanations as good as or better than a centralized model, while extreme heterogeneity causes explanations to become spatially incoherent and causally misleading. The authors provide practical guidance for determining when federated deepfake detection can be safely deployed and when its explanations lose trustworthiness.
By Muhammad Uzair Shahid, Muhammad Zabil Mehboob
Large Language Models (LLMs) treat all context tokens uniformly, which creates prompt injection vulnerabilities and unreliable instruction execution. The paper introduces LLMON (LLM Object Notation), a post‑training framework that separates executable instructions from passive data and requires explicit identifier‑based execution. Using a 4.6 B‑token structured corpus and post‑training on Granite‑4.1 models, LLMON reduces indirect prompt‑injection success rates from 53–100 % to 0.02–5.6 % while maintaining general capabilities, and provides attention‑based interpretability metrics to diagnose failure modes.
By Basel Shbita, Michael Hind, Chad DeLuca, Nathan Fulton, Shubhi Asthana, Farhan Ahmed, Anisa Halimi, Yada Zhu, Mauro Martino, David Cox, Dan Gutfreund
The paper introduces CHORD, a relational graph discrimination framework designed to detect hallucinations across different large language models (LLMs). It addresses the challenge of cross-model representation heterogeneity by modeling hidden states as joint relational graphs over layers, tokens, and features, and employs relation-aware attention and meta-learning to extract transferable detection signals. Experiments demonstrate that CHORD outperforms existing methods in cross-model generalization.
By Yongxin Deng (University of Technology Sydney), Zhen Fang (University of Technology Sydney), Guansong Pang (Singapore Management University), Sharon Li (University of Wisconsin-Madison), Ling Chen (University of Technology Sydney)
The paper investigates how self‑improving large language model agents can inadvertently remove redundant safeguards through acceptance gates that evaluate edits based on task success alone. By systematically deleting safeguards in two agent environments and running 411 improvement loops under various gate rules, the authors show that when one safeguard is removed, others still block harm, yet the gate may still delete it if it improves task performance. Larger evaluation batches and certain optimization strategies (e.g., GEPA) can lead to the loss of multiple safeguards, exposing the system to harm when the harness is reused with a weaker model.
By Hau-Sing So, Daixian Liu, Weicai Long
The paper introduces Value Probes, a method that uses linear probing on a language model’s activations to estimate a value function for future rewards in automated red‑team auditing. By treating the audit as a Markov decision process and applying beam search guided by these value estimates, the authors achieve a roughly 50% increase in audit success across 18 target behaviors and three model scales. The approach also demonstrates that Value Probes can be trained on off‑policy data from smaller models, offering a more efficient alternative to costly language‑model follow‑ups.
By Jonathan F. Jao, Andrew Kyle Lampinen
The paper examines how AI‑written research papers often contain misattributed numerical claims. By re‑analyzing 540 judgments from a benchmark of nine AI agents on fifteen research problems, the authors manually labeled 256 claims and found that most failures stem from misalignment between the paper and the underlying work rather than outright cheating. They propose a deterministic claim‑to‑evidence checker that achieves over 90% precision and, when combined with an LLM verifier, offers a more reliable audit of AI‑generated research.
By Mosab Alfaqeeh, Omar Hourani, Muath Hourani, Hashem Alfaqeeh
The paper introduces LLM-Integrated Bayesian State Space Models (LBS) for multimodal time-series forecasting, combining a state space model backbone with a pretrained large language model to handle both numerical and textual data. LBS allows flexible lookback and forecast windows, principled uncertainty quantification, and improved temporal generalization. Experiments on the TextTimeCorpus benchmark show a 13.20% performance gain over the previous state‑of‑the‑art while providing human‑readable forecast summaries.
By Sungjun Cho, Changho Shin, Suenggwan Jo, Xinya Yan, Shourjo Aditya Chaudhuri, Frederic Sala
The article introduces Large Language Models (LLMs) to scientists and professionals, covering core concepts such as tokenisation, embeddings, training, inference, context windows, prompting, and instruction hierarchies without delving into mathematics. It emphasizes the limitations of LLM outputs—hallucinations, uncertainty, reproducibility, bias—and stresses the importance of systematic verification, human oversight, and responsible handling of sensitive data. Practical exercises guide readers through running a local LLM with Ollama and Open WebUI, focusing on evidence retrieval, source verification, and distinguishing plausible language from scientifically supported conclusions.
By Umberto Michelucci
DuQUAD is a dual‑view framework for multi‑document question answering that addresses the "Lost in the Middle" problem by combining a Local Agent guided by a Structural Fusion Score with a Global Agent that reasons over the entire document set. It further prunes evidence using explicit quality scoring and sentence‑level refinement to reduce noise. The approach outperforms existing mitigation baselines, improving accuracy by an average of 3.0% and gold evidence recall by 9.4%, while cutting computational cost.
By Yeonjun Choi, Seyeon Jeong, Jong Wook Kim, Beakcheol Jang
The paper investigates how small entry‑wise errors in machine‑learned Kohn‑Sham Hamiltonians can still lead to inaccurate eigenvalues and eigenvectors. It shows that existing physics‑informed losses only partially capture the error, and proposes using the cumulative local density of states (cumulative LDOS) as a complete training signal that uniquely determines the Hamiltonian. Fine‑tuning three Hamiltonian networks with this loss, called FermiScope, reduces HOMO‑LUMO gap errors without increasing element‑wise error, and the improvement generalizes to out‑of‑distribution and molecular‑dynamics scenarios.
By Fang Sun, Hekai Cheng, Ruoyan Li, Yuanzhou Chen, Wei Wang, Jason Cong, Yizhou Sun
The paper introduces weighted t‑SNE, a variant of t‑SNE that incorporates feature importance scores from feature scorers into the 2‑D embedding of high‑dimensional data. It explains how to implement the method, proposes using the silhouette coefficient to evaluate the resulting visualizations, and demonstrates its use on synthetic and real‑world tabular datasets with nine different feature scorers. The approach allows practitioners to compare feature scorers visually and to interpret how machine‑learning models organize data based on learned feature importance.
By Bruno Iochins Grisci, Mario Inostroza-Ponta, M\'arcio Dorn
The article "Hybrid AI Patterns: Thirteen Ways to Couple Language Models with Rules, and How to Choose Among Them" presents a catalogue of thirteen design patterns for integrating language models with rule-based systems. It details each pattern’s intent, motivation, applicability, participants, collaborations, consequences, implementation, known uses, and related patterns, and identifies five invariants and five failure patterns that apply to all combinations. A decision guide is provided to help select an appropriate pattern based on problem context, supported by measurements of performance and design costs.
By Siegfried Handschuh, G\"otz-Henrik Wiegand, Lorena Raichle