Isolater - Feed

HN delfrrr 4/16/2026

Charts-of-Thought: Enhancing LLM Visualization Literacy (PDF, 2026)

Research paper on Charts-of-Thought method for enhancing LLM visualization literacy and understanding of data representations.

Ax Arun Sharma 4/16/2026

Spatial Atlas: Compute-Grounded Reasoning for Spatial-Aware Research Agent Benchmarks

Spatial Atlas framework instantiating compute-grounded reasoning paradigm for spatial-aware research agents with A2A architecture.

Ax Hao Yan, Yuliang Liu, Xingchen Liu, Yuyi Zhang, Minghui Liao, Jihao Wu, Wei Chen, Xiang Bai 4/16/2026

DocSeeker: Structured Visual Reasoning with Evidence Grounding for Long Document Understanding

DocSeeker multimodal LLM system for long document understanding with evidence grounding, addressing signal-to-noise and weak supervision challenges.

Ax Yaxuan Li, Yuxin Zuo, Bingxiang He, Jinqian Zhang, Chaojun Xiao, Cheng Qian, Tianyu Yu, Huan-ang Gao, Wenkai Yang, Zhiyuan Liu, Ning Ding 4/16/2026

Rethinking On-Policy Distillation of Large Language Models: Phenomenology, Mechanism, and Recipe

Systematic investigation of on-policy distillation dynamics in LLM post-training, identifying conditions for success and failure mechanisms.

Ax Mingxiang Tao, Yu Tian, Wenxuan Tu, Yue Yang, Xue Yang, Xiangyan Tang 4/16/2026

Safe-FedLLM: Delving into the Safety of Federated Large Language Models

Security analysis of federated learning for LLMs, investigating attack surfaces and defenses against malicious clients in open environments.

Ax Alejandro Carrasco, Mariko Storey-Matsutani, Victor Rodriguez-Fernandez, Richard Linares 4/16/2026

GUIDE: Guided Updates for In-context Decision Evolution in LLM-Driven Spacecraft Operations

GUIDE framework for LLM-driven spacecraft operations using in-context learning to improve agent decisions across episodes without weight updates.

Ax Han Li, Yifan Yao, Letian Zhu, Rili Feng, Hongyi Ye, Jiaming Wang, Yancheng He, Pengyu Zou, Lehan Zhang, Xinping Lei, Haoyang Huang, Ken Deng, Ming Sun, Zhaoxiang Zhang, He Ye, Jiaheng Liu 4/16/2026

CodeTracer: Towards Traceable Agent States

CodeTracer framework for debugging and tracing AI agent state transitions, error propagation, and tool orchestration in code agents.

Ax Rajat Khanda, Mohammad Baqar Sambuddha Chakrabarti, Satyasaran Changdar 4/16/2026

Adaptive Memory Crystallization for Autonomous AI Agent Learning in Dynamic Environments

Adaptive Memory Crystallization architecture enabling continual learning in autonomous AI agents without catastrophic forgetting.

Ax Fei Ding, Yongkang Zhang, youwei wang, Zijian Zeng 4/16/2026

Design Conditions for Intra-Group Learning of Sequence-Level Rewards: Token Gradient Cancellation

Analysis of sequence-level reward learning in reinforcement learning for reasoning models, addressing gradient cancellation and learning efficiency.

Ax Saumya Goyal, Rohith Rongali, Ritabrata Ray, Barnab\'as P\'oczos 4/16/2026

Generalization Guarantees on Data-Driven Tuning of Gradient Descent with Langevin Updates

Langevin Gradient Descent algorithm with generalization guarantees for hyperparameter tuning in convex regression via learning to learn.

Ax Wenqing Li, Xu Feng, Peixue Jiang, Yinhai Zhu 4/16/2026

Automated co-design of high-performance thermodynamic cycles via graph-based hierarchical reinforcement learning

Graph-based hierarchical reinforcement learning for automated co-design of thermodynamic cycle parameters.

Ax Aadyot Bhatnagar, Peter M{\o}rch Groth, Ali Madani 4/16/2026

Pareto-Optimal Offline Reinforcement Learning via Smooth Tchebysheff Scalarization

Pareto-optimal offline reinforcement learning via Tchebysheff scalarization for multi-objective LLM alignment and optimization.

Ax Chuangtao Chen, Grace Li Zhang, Xunzhao Yin, Cheng Zhuo, Bing Li, Ulf Schlichtmann 4/16/2026

KV Packet: Recomputation-Free Context-Independent KV Caching for LLMs

KV Packet enables context-independent key-value caching for LLMs without recomputation, improving inference latency.

Ax Iv\'an Olarte Rodr\'iguez, Anja Jankovic, Thomas B\"ack, Elena Raponi 4/16/2026

Does Dimensionality Reduction via Random Projections Preserve Landscape Features?

Studies whether dimensionality reduction via random projections preserves landscape features for exploratory landscape analysis.

Ax Kevin Wilkinghoff, Neelu Madan, Juan Miguel Valverde, Kamal Nasrollahi, Radu Tudor Ionescu, Rafal Wisniewski, Thomas B. Moeslund, Wenwu Wang, Zheng-Hua Tan 4/16/2026

Out of Context: Reliability in Multimodal Anomaly Detection Requires Contextual Inference

Context-dependent anomaly detection framework for multimodal data recognizing that anomalies depend on contextual factors.

Ax Ankit Lade, Sai Krishna J., Indar Kumar 4/16/2026

Bias-Corrected Adaptive Conformal Inference for Multi-Horizon Time Series Forecasting

Bias-corrected adaptive conformal inference for multi-horizon time series forecasting with distribution shift adaptation.

Ax Sanjar Khudoyberdiev, Arman Bekov 4/16/2026

Counterfactual Peptide Editing for Causal TCR--pMHC Binding Inference

Counterfactual invariant prediction framework prevents shortcut learning in TCR-pMHC binding neural prediction models.

Ax Yilang Zhang, Abraham Jaeger Mountain, Bingcong Li, Georgios B. Giannakis 4/16/2026

Binomial Gradient-Based Meta-Learning for Enhanced Meta-Gradient Estimation

Binomial gradient-based meta-learning approach to reduce computational overhead in gradient-based meta-learning.

Ax Anton Saenko, Pranshav Gajjar, Abiodun Ganiyu, Vijay K. Shah 4/16/2026

Enhancing Confidence Estimation in Telco LLMs via Twin-Pass CoT-Ensembling

Twin-pass chain-of-thought ensembling method to improve confidence estimation reliability in telecommunications LLMs.

Ax Gabriel Afriat, Xiang Meng, Shibal Ibrahim, Hussein Hazimeh, Rahul Mazumder 4/16/2026

MOONSHOT : A Framework for Multi-Objective Pruning of Vision and Large Language Models

MOONSHOT framework for multi-objective one-shot pruning of vision and large language models without retraining.

Ax Samrendra Roy, Souvik Chakraborty, Syed Bahauddin Alam 4/16/2026

Beyond Uniform Sampling: Synergistic Active Learning and Input Denoising for Robust Neural Operators

Combines active learning and input denoising to improve robustness of neural operators against adversarial perturbations.

Ax Jiahao Shao, Anam Nawaz Khan, Christopher Brett, Tom Berg, Xueping Li, Bing Yao 4/16/2026

Multi-Task LLM with LoRA Fine-Tuning for Automated Cancer Staging and Biomarker Extraction

Multi-task LLM framework with LoRA fine-tuning for automated cancer staging and biomarker extraction from pathology reports.

Ax Mohsen Nayebi Kerdabadi, Arya Hadizadeh Moghaddam, Chen Chen, Dongjie Wang, Zijun Yao 4/16/2026

Text-Attributed Knowledge Graph Enrichment with Large Language Models for Medical Concept Representation

Uses LLMs to enrich knowledge graphs for medical concept representation in EHR mining and clinical prediction tasks.

Ax Jingyun Jia, Chandan Singh, Rich Caruana, Ben Lengerich 4/16/2026

Selecting Feature Interactions for Generalized Additive Models by Distilling Foundation Models

TabDistill method leverages tabular foundation models to identify feature interactions for generalized additive models on tabular data.

Ax Yiping Li, Zhiyu An, Wan Du 4/16/2026

When Less Latent Leads to Better Relay: Information-Preserving Compression for Latent Multi-Agent LLM Collaboration

Orthogonal Backfill compression for LLM multi-agent systems reducing KV cache relay costs while preserving communication context and information.

Ax Run Wang, Victor J. B. Jung, Philip Wiese, Sebastian Frey, Giusy Spacone, Francesco Conti, Alessio Burrello, Luca Benin 4/16/2026

BioTrain: Sub-MB, Sub-50mW On-Device Fine-Tuning for Edge-AI on Biosignals

BioTrain framework enabling sub-50mW on-device fine-tuning for MCU-based wearable edge AI on biosignals addressing domain shift and privacy.

Ax Angelo Moroncelli, Matteo Rufolo, Gunes Cagin Aydin, Asad Ali Shahid, Loris Roveda 4/16/2026

Diffusion Sequence Models for Generative In-Context Meta-Learning of Robot Dynamics

Diffusion sequence models and Transformer meta-models for in-context robot dynamics learning addressing distributional shifts and real-time constraints.

Ax Zhengyu Fang, Zhimeng Jiang, Huiyuan Chen, Xiaoge Zhang, Tianyi Li, Kaiyu Tang, Xiao Li, Jing Li 4/16/2026

Dataset-Level Metrics Attenuate Non-Determinism: A Fine-Grained Non-Determinism Evaluation in Diffusion Language Models

Fine-grained non-determinism evaluation in diffusion language models showing dataset-level metrics mask run-to-run variations and condition sensitivity.

Ax Xingjian Zhao, Mohammad Mohammadi Amiri, Malik Magdon-Ismail 4/16/2026

WIN-U: Woodbury-Informed Newton-Unlearning as a retain-free Machine Unlearning Framework

WIN-U: Woodbury-informed Newton method for machine unlearning in LLMs enabling 'right to be forgotten' without requiring retain set data.

Ax Jason Kong, Nilesh Prasad Pandey, Flavio Ponzina, Tajana Rosing 4/16/2026

A KL Lens on Quantization: Fast, Forward-Only Sensitivity for Mixed-Precision SSM-Transformer Models

Forward-only KL-based sensitivity analysis for mixed-precision quantization of hybrid SSM-Transformer LLMs targeting edge device deployment.

Ax Shentong Mo 4/16/2026

Chain of Uncertain Rewards with Large Language Models for Reinforcement Learning

Proposes Chain of Uncertain Rewards method for designing reward functions in RL using LLMs, addressing inefficiencies in manual reward design and capturing intermediate uncertainties.

Ax Xiaole Su, Kasey Zhang, Andy Lyu 4/16/2026

SFT-GRPO Data Overlap as a Post-Training Hyperparameter for Autoformalization

Research on SFT-GRPO data overlap as a post-training hyperparameter for Lean 4 autoformalization using Qwen3-8B, ablating overlap percentages (0%, 30%, 100%).

Ax Jing Sun 4/16/2026

Representation over Routing: Overcoming Surrogate Hacking in Multi-Timescale PPO

Analysis of multi-timescale PPO revealing surrogate hacking when fusing multi-scale signals, proposing representation-focused solutions.

Ax Mintu Dutta, Ritesh Vyas, Mohendra Roy 4/16/2026

From Alignment to Prediction: A Study of Self-Supervised Learning and Predictive Representation Learning

Study of self-supervised learning and predictive representation learning comparing alignment and reconstruction approaches.

Ax Kenji Kubo, Shunsuke Kamiya, Masanori Koyama, Kohei Hayashi, Yusuke Iwasawa, Yutaka Matsuo 4/16/2026

C-voting: Confidence-Based Test-Time Voting without Explicit Energy Functions

Confidence-based test-time voting mechanism for latent recurrent neural networks enabling test-time scaling without explicit energy functions.

Ax Yongil Choi 4/16/2026

Learning Inference Concurrency in DynamicGate MLP Structural and Mathematical Justification

DynamicGate MLP architecture permitting concurrent learning and inference by separating routing from representation parameters.

Ax Hevish Cowlessur, Chandra Thapa, Tansu Alpcan, Seyit Camtepe 4/16/2026

Parameter-efficient Quantum Multi-task Learning

Parameter-efficient quantum multi-task learning with shared backbone and task-specific heads for quantum neural networks.

Ax Qin Zhou, Guoyan Liang, Qianyi Yang, Jingyuan Chen, Sai Wu, Chang Yao, Zhe Wang 4/16/2026

Enhancing Reinforcement Learning for Radiology Report Generation with Evidence-aware Rewards and Self-correcting Preference Learning

RL approach for radiology report generation using evidence-aware rewards and self-correcting preference learning for clinical alignment.

Ax Xiaohua Wang, Muzhao Tian, Yuqi Zeng, Zisu Huang, Jiakang Yuan, Bowen Chen, Jingwen Xu, Mingbo Zhou, Wenhao Liu, Muling Wu, Zhengkang Guo, Qi Qian, Yifei Wang, Feiran Zhang, Ruicheng Yin, Shihan Dou, Changze Lv, Tao Chen, Kaitao Song, Xu Tan, Tao Gui, Xiaoqing Zheng, Xuanjing Huang 4/16/2026

Reward Hacking in the Era of Large Models: Mechanisms, Emergent Misalignment, Challenges

Analyzes reward hacking vulnerabilities in RLHF and alignment approaches for LLMs, examining mechanisms and emergent misalignment issues.

Ax Aram Ebtekar, Michael K. Cohen 4/16/2026

Golden Handcuffs make safer AI agents

Bayesian mitigation strategy for safer AI agents using expanded subjective reward range to prevent reward hacking via risk aversion.

Ax Mark Rofin, Aditya Varre, Nicolas Flammarion 4/16/2026

(How) Learning Rates Regulate Catastrophic Overtraining

Studies how learning rates regulate catastrophic overtraining in LLM fine-tuning through catastrophic forgetting lens.

Ax Yinsong Chen, Samson S. Yu, Kashem M. Muttaqi 4/16/2026

A Bayesian Framework for Uncertainty-Aware Explanations in Power Quality Disturbance Classification

Bayesian framework for uncertainty-aware explainable AI in power quality disturbance classification with instance-specific interpretations.

Ax Thomas Bartz-Beielstein 4/16/2026

Optimization with SpotOptim

Python package for surrogate-model-based optimization using Kriging, Expected Improvement, and multi-objective extensions for expensive functions.

Ax Angelo Moroncelli, Roberto Zanetti, Marco Maccarini, Loris Roveda 4/16/2026

Jump-Start Reinforcement Learning with Vision-Language-Action Regularization

Combines Vision-Language-Action models with RL for robotic manipulation, enabling efficient long-horizon task learning with sparse rewards.

Ax Pranav Mahajan, Ben Seymour 4/16/2026

Soft $Q(\lambda)$: A multi-step off-policy method for entropy regularised reinforcement learning using eligibility traces

Multi-step off-policy soft Q-learning using eligibility traces for entropy-regularized RL with formal n-step formulation.

Ax Dongjie Fu, Fangming Feng, Xize Cheng, Linjun Li, Zhou Zhao, Tao Jin 4/16/2026