The paper introduces REPAIR, a method that corrects lossy user preference states in personalization encoders by comparing cached representations with the current state in a learned coordinate space. REPAIR selectively aggregates corrective evidence from past interactions and adds it to the state before the task head, enabling encoder–host repair without re‑encoding history. Experiments on MovieLens, PENS, MIND, and Amazon Reviews 2023 show that training only REPAIR improves MRR and nDCG@10 across all twelve recommendation hosts, while head‑only finetuning yields smaller gains.
By Parthiv Chatterjee, Dhiraj Golhar, Ummesalma Diwan, Sourish Dasgupta, Manjunath Joshi, Tanmoy Chakraborty
The paper introduces IDiom, an autoregressive protein language model trained on a large dataset of intrinsically disordered protein regions (IDRs) from AlphaFold, and demonstrates that it can generate sequences matching natural IDR composition, motifs, and disorder. It further presents RL‑SAE, a reinforcement learning approach that uses sparse autoencoder features to steer generation toward specific functional patterns, achieving high activation of targeted features and improved predicted subcellular localization and transcriptional activity. The combination of IDiom and RL‑SAE allows interpretable, composable IDR design by explicitly controlling function‑associated sequence features.
By Jason X. Liu, Sebastian Ibarraran, Frank Hu, Soojung Yang, Xinyu A. Feng, Abigail Park, Anagha Aneesh, Lacramioara Bintu, Alexander R. Dunn, Grant M. Rotskoff
The paper evaluates System One decision models—typed models that output probabilities for branching decisions—against supervised classifiers and generative language models on automated decision gate tasks. Eight checkpoints from six families, including the hosted model Jev, were benchmarked on workflow, intent, and social‑science items, showing that small trained classifiers match or slightly outperform decision models on intent and workflow when labels are available, while decision models outperform zero‑shot classifiers when labels are absent. The study also explores calibration, risk thresholds, and cost‑efficiency trade‑offs, providing condition‑dependent design guidelines for automated decision gates.
By Amir Rafe, Subasish Das
The paper presents a system for the IEEE BigData 2026 Cup on Explainable Suicide Risk Assessment on Social Media. It tackles three tasks—risk-level classification, evidence phrase extraction, and multi-label factor identification—using Qwen2.5-Instruct models adapted with quantized low-rank adaptation (QLoRA) and an answer-masked causal language-model objective. The final system achieved a composite score of 0.7738, with 0.8089 on Task 1 and 0.6919 on Task 2, demonstrating that task‑specific training and tailored aggregation improve performance across the three tasks.
By Xuan Zhong Feng, Geoffrey Martin, Hexin Dong, Yifan Peng
Propaganda detection is an essential task in natural language processing (NLP), particularly in the context of manipulative political communications. However, identifying specific propaganda technique...
The paper introduces TESS, a scalable data‑selection framework that replaces per‑sample weights with a selection network to improve transferability across datasets and model sizes. It identifies instability in existing meta‑learning for training‑data selection (MTS) due to weight suppression and overreliance on easy features, and proposes a Pointwise Value Matching objective to address these issues. Experiments on large language model safety and instruction tuning show strong transfer from subsets to full corpora and from smaller to larger models.
By Zilin Du, Bowen Yang, Boyang Albert Li
The paper introduces an accelerated algorithm for sparse regularized partial optimal transport (POT). It reformulates the problem using penalty-based techniques that allow efficient gradient updates while maintaining the original structure, and supports a wide range of regularizers that encourage structured sparsity. Experiments on color transfer, domain adaptation, and point cloud registration show the method outperforms existing baselines in transport cost, sparsity, and convergence speed.
By Khoa Nguyen, Dung T. Nguyen, Thong Huynh, Hoang-Hiep Nguyen-Mau, Anh Nguyen, Minh Ngoc Dinh, Juho Kannala
The paper proposes ANCHOR, a composition policy that decouples crystal structure prediction (CSP) from de novo crystal generation (DNG) by using a frozen CSP model as a fixed evaluator and training the policy with multi‑objective rewards, including adaptive novelty. ANCHOR significantly improves metrics such as MSUN and SUN compared to traditional DNG approaches, and demonstrates transferability across different CSP backbones and distillation into Crystalite‑CSP. The study shows that fine‑tuning DNG models directly on these rewards yields limited gains, whereas the ANCHOR framework better leverages the CSP prior for more efficient crystal discovery.
By Emma Lei Hovmand, Jonas Elsborg, Melih Kandemir, Arghya Bhowmik
The paper introduces Concept-Driven Domain Adaptation (CDDA), a three-stage framework that adapts vision‑language models for concept‑to‑example video retrieval in educational settings. CDDA first structures textual embeddings using textbook and teacher‑handbook concept pairs, then transfers this geometry to documentary visuals with a frozen visual encoder, and finally jointly fine‑tunes both encoders with sparse visual concept supervision. On a middle‑school physics benchmark, CDDA outperforms several multimodal baselines in retrieving concept‑driven moments while preserving concrete image‑text alignment.
By Haiming Zhao, Tai Wang, Kun Zhang, Xicheng Peng, Zhiyang Li
The paper investigates whether the sparsity of Mixture-of-Experts (MoE) models leads to intrinsic semantic organization across modalities and domains. It shows that experts naturally specialize semantically even without explicit modular training. The authors propose ExpertLens, a data‑free method that decodes router weights to identify domain‑specialized experts, enabling selective fine‑tuning that matches or exceeds full fine‑tuning while updating only 21.7–47.0% of parameters and achieving a 4.0× speedup, outperforming LoRA in both performance and efficiency.
By Damiano Marsili, Raphi Kang, Aditya Mehta, Pietro Perona, Georgia Gkioxari
The paper introduces PrecepTron, a 32‑billion‑parameter language model fine‑tuned with low‑rank adaptation to evaluate clinical reasoning in large language models (LLMs) at a physician level. It also releases GRAND‑ROUNDS, a benchmark of 9,217 scored responses from 160 clinicians across seven studies. Using PrecepTron, the authors replicate key findings from major medical AI studies and explore new questions about LLM diagnostic accuracy, demonstrating that fine‑tuned models can provide consistent, scalable physician‑level scoring.
By Thomas A. Buckley, Zahir Kanjee, Peter G. Brodeur, Byron Crowe, Anthony M. Pettinato, Aashna P. Shah, Adrian D. Haimovich, Liam G. McCoy, Daniel Restrepo, Jason A. Freed, Ethan Goh, Jonathan H. Chen, Laura Zwaan, Katherine E. Goodman, Daniel J. Morgan, Raja-Elie E. Abdulnour, Adam Rodman, Arjun K. Manrai
The paper presents a hierarchical reinforcement learning framework for jointly optimizing routing and switch placement in chip design. It builds solutions from a minimal routing graph using three operations—switch expansion, switch placement, and route refinement—that maintain routing validity by construction. Neural‑guided Gumbel Monte Carlo Tree Search and pretraining across floorplans are shown to improve solution quality over non‑learning methods.
By Dorian Gailhard, Ugo Lecerf, Enzo Tartaglione, Donatello Conte, Jhony H. Giraldo
The paper introduces a reinforcement learning-guided framework for graph transformations aimed at optimizing sparse triangular solve (SpTRSV) kernels. By treating graph transformation as a sequential decision-making task, the RL agent learns matrix-dependent policies that reduce the number of levels by 23% and the coefficient of variation of level costs by 29%, while modifying less than 1% of matrix rows. Experimental results on real-world sparse matrices show significant performance improvements and demonstrate that the learned policies can transfer to unseen matrices via curriculum learning and fine-tuning.
By Buse Y{\i}lmaz
The paper introduces AEGIS, a buffer‑free, layer‑wise orthogonal gradient projection framework that enables continuous fine‑tuning of Vision‑Language Models for robotic manipulation while protecting pre‑trained visual reasoning. AEGIS uses pre‑training activation statistics as static anchors and applies a Wasserstein‑2 transport penalty to generate an anchor‑restoration gradient, followed by a dual‑backward pass that orthogonalizes task gradients against this restoration vector. Experiments on the LIBERO manipulation benchmark with PaliGemma2‑3B show that AEGIS preserves Visual Question Answering performance and baseline loss while achieving continuous action convergence, all without replay buffers, teacher models, or co‑training data.
By Guransh Singh
The paper "Where's Waldo? Query-language Preference under Cross-lingual Knowledge Disparities" introduces the Waldo benchmark, a multilingual QA dataset built from Wikipedia that focuses on knowledge gaps and conflicts across languages. It evaluates eight models in five languages and finds that when a fact is missing in one language, models tend to use evidence from the other language, but when conflicting accounts exist, responses align strongly with the query language, leading to different answers for semantically identical questions. The study also explores mitigation strategies, including ablating attention heads and LoRA-based training, which can reduce the preference gap by up to 61.5%.
By Dayeon Ki, Ruochen Zhang, Silviu Cucerzan, Ryen W. White, Ning Gao
arXiv:2610. 01493v1 Announce Type: cross Abstract: Iterative fine-tuning on synthetic data causes \emph{model collapse}: output diversity narrows as rare patterns are progressively lost, a signature most visible as phrase-level repetition.
By Lewis Mitchell
Generative Cinematographer (GenCine) lifts a single image into an editable 3D scene scaffold, allowing artists to jointly author camera paths and foreground motion using local 3D motion handles. The system projects these controls into guidance maps that encode handle positions and colors across frames, enabling a pretrained video model to follow camera-relative, piecewise-rigid motion without physics simulation. Training recovers controls from real videos and synthetic ground-truth geometry, and experiments demonstrate consistent camera-relative motion, improved geometric consistency, and strong controllability across diverse real-world scenes.
By Jiahan Zhang, Chaohao Yang, Namitha Guruprasad, Vivekjyoti Banerjee, Trong-Tung Nguyen, Alan Yuille, Anand Bhattad
REALM is a retrospective knowledge distillation framework that enables causal decoding of behavior from local field potentials (LFPs). It trains a bidirectional Mamba‑2 teacher on multi‑session data using continuous masked autoencoding, then distills its representations into a compact causal student model. The resulting LFP‑only decoder achieves the highest mean accuracy among compared methods, surpassing state‑of‑the‑art baselines while using fewer parameters and less pretraining time.
By Peicheng Wu, Zhenyu Bu, Runze Ma, Lin Du
GPart introduces a new parameter‑efficient fine‑tuning technique that directly maps a low‑dimensional trainable vector into the full weight space using a sparse, isometric partition matrix. Unlike LoRA, GPart eliminates the bilinear reconstruction step, preserving exact end‑to‑end isometry and reducing the checkpoint to just the vector and a random seed. Experiments across NLP, vision, and reasoning tasks show that GPart matches or surpasses existing PEFT methods while using far fewer parameters and offering a simpler, more tractable parameterization.
By Paolo Mandica, Micha{\l} Brzozowski, Zuzanna Dubanowska, Neo Christopher Chung
The study investigates whether the organization of mathematical training data for large language models (LLMs) should be based on topic or on reasoning approach. Using two counterbalanced experimental designs, the authors compare fine‑tuned models trained on problems that share the same solution method (same‑approach, SA) versus those that share the same topic (same‑topic, ST). Across five base models and multiple training seeds, SA consistently outperforms ST, with mean gains of 10.8–14.3 percentage points, even though ST sources are more similar to targets in embedding and lexical metrics.
By Sajad Goudarzi, Samaneh Zamanifard, Seyed Amin Seyed Haeri, Moloud Nasiri, Hamed Rahimian