CUEBES

Can David Beat Goliath? On Multi-Hop Reasoning with Resource-Constrained Agents

arXiv:2601.21699v1 Announce Type: new Abstract: While reinforcement learning (RL) has empowered multi-turn reasoning agents with retrieval and tools, existing successes largely depend on extensive on-policy...

Policy Energy

arXiv CS Jan 30

Toward Culturally Aligned LLMs through Ontology-Guided Multi-Agent Reasoning

arXiv:2601.21700v1 Announce Type: new Abstract: Large Language Models (LLMs) increasingly support culturally sensitive decision making, yet often exhibit misalignment due to skewed pretraining data and...

World News Policy

arXiv CS Jan 30

Age Aware Content Fetching and Broadcast in a Sensing-as-a-Service System

arXiv:2601.21701v1 Announce Type: new Abstract: We consider a Sensing-as-a-Service (S2aaS) system consisting of a sensor, a set of users, and a sensor cloud service provider...

Policy Mathematics

arXiv CS Jan 30

Beyond Forgetting: Machine Unlearning Elicits Controllable Side Behaviors and Capabilities

arXiv:2601.21702v1 Announce Type: new Abstract: We consider representation misdirection (RM), a class of LLM unlearning methods that achieves forgetting by manipulating the forget-representations, that is,...

Psychology Software

arXiv CS Jan 30

SmartMeterFM: Unifying Smart Meter Data Generative Tasks Using Flow Matching Models

arXiv:2601.21706v1 Announce Type: new Abstract: Smart meter data is the foundation for planning and operating the distribution network. Unfortunately, such data are not always available...

Policy Software

arXiv CS Jan 30

Adaptive Kernel Methods

arXiv:2601.21707v1 Announce Type: new Abstract: Kernel methods approximate nonlinear maps in a data-driven manner by projecting the target map onto a finite-dimensional Hilbert space called...

Biology Software

arXiv CS Jan 30

FBS: Modeling Native Parallel Reading inside a Transformer

arXiv:2601.21708v1 Announce Type: new Abstract: Large language models (LLMs) excel across many tasks, yet inference is still dominated by strictly token-by-token autoregression. Existing acceleration methods...

Energy Policy

arXiv CS Jan 30

Why Attention Patterns Exist: A Unifying Temporal Perspective Analysis

arXiv:2601.21709v1 Announce Type: new Abstract: Attention patterns play a crucial role in both training and inference of large language models (LLMs). Prior works have identified...

Software Psychology

arXiv CS Jan 30

TACLer: Tailored Curriculum Reinforcement Learning for Efficient Reasoning

arXiv:2601.21711v1 Announce Type: new Abstract: Large Language Models (LLMs) have shown remarkable performance on complex reasoning tasks, especially when equipped with long chain-of-thought (CoT) reasoning....

Policy Artificial Intelligence

arXiv CS Jan 30

CoFreeVLA: Collision-Free Dual-Arm Manipulation via Vision-Language-Action Model and Risk Estimation

arXiv:2601.21712v1 Announce Type: new Abstract: Vision Language Action (VLA) models enable instruction following manipulation, yet dualarm deployment remains unsafe due to under modeled selfcollisions between...

Policy Robotics

arXiv CS Jan 30

Disentangling perception and reasoning for improving data efficiency in learning cloth manipulation without demonstrations

arXiv:2601.21713v1 Announce Type: new Abstract: Cloth manipulation is a ubiquitous task in everyday life, but it remains an open challenge for robotics. The difficulties in...

Robotics Environment

arXiv CS Jan 30

E-mem: Multi-agent based Episodic Context Reconstruction for LLM Agent Memory

arXiv:2601.21714v1 Announce Type: new Abstract: The evolution of Large Language Model (LLM) agents towards System~2 reasoning, characterized by deliberative, high-precision problem-solving, requires maintaining rigorous logical...

Biology Energy

arXiv CS Jan 30

DreamActor-M2: Universal Character Image Animation via Spatiotemporal In-Context Learning

arXiv:2601.21716v1 Announce Type: new Abstract: Character image animation aims to synthesize high-fidelity videos by transferring motion from a driving sequence to a static reference image....

Psychology Chemistry

arXiv CS Jan 30

When does predictive inverse dynamics outperform behavior cloning?

arXiv:2601.21718v1 Announce Type: new Abstract: Behavior cloning (BC) is a practical offline imitation learning method, but it often fails when expert demonstrations are limited. Recent...

Environment Psychology

arXiv CS Jan 30

LoRA and Privacy: When Random Projections Help (and When They Don't)

arXiv:2601.21719v1 Announce Type: new Abstract: We introduce the (Wishart) projection mechanism, a randomized map of the form $S \mapsto M f(S)$ with $M \sim W_d(1/r...

Cybersecurity Policy

arXiv CS Jan 30

Enhancing Language Models for Robust Greenwashing Detection

arXiv:2601.21722v1 Announce Type: new Abstract: Sustainability reports are critical for ESG assessment, yet greenwashing and vague claims often undermine their reliability. Existing NLP models lack...

Sustainability Environment

arXiv CS Jan 30

Procedural Pretraining: Warming Up Language Models with Abstract Data

arXiv:2601.21725v1 Announce Type: new Abstract: Pretraining directly on web-scale corpora is the de facto paradigm for building language models. We study an alternative setting where...

Artificial Intelligence Biology

arXiv CS Jan 30

DropoutTS: Sample-Adaptive Dropout for Robust Time Series Forecasting

arXiv:2601.21726v1 Announce Type: new Abstract: Deep time series models are vulnerable to noisy data ubiquitous in real-world applications. Existing robustness strategies either prune data or...

Software World News

arXiv CS Jan 30

Amortized Spectral Kernel Discovery via Prior-Data Fitted Network

arXiv:2601.21731v1 Announce Type: new Abstract: Prior-Data Fitted Networks (PFNs) enable efficient amortized inference but lack transparent access to their learned priors and kernels. This opacity...

Software Mathematics

arXiv CS Jan 30

CE-GOCD: Central Entity-Guided Graph Optimization for Community Detection to Augment LLM Scientific Question Answering

arXiv:2601.21733v1 Announce Type: new Abstract: Large Language Models (LLMs) are increasingly used for question answering over scientific research papers. Existing retrieval augmentation methods often rely...

Software Policy

arXiv CS Jan 30

A reduced basis method for parabolic PDEs based on a space-time least squares formulation

arXiv:2601.21736v1 Announce Type: new Abstract: In this work, we present a POD-greedy reduced basis method for parabolic partial differential equations (PDEs), based on the least...

Technology Software

arXiv CS Jan 30

Mixed-Precision Training and Compilation for RRAM-based Computing-in-Memory Accelerators

arXiv:2601.21737v1 Announce Type: new Abstract: Computing-in-Memory (CIM) accelerators are a promising solution for accelerating Machine Learning (ML) workloads, as they perform Matrix-Vector Multiplications (MVMs) on...

Software Policy

arXiv CS Jan 30

From Global to Granular: Revealing IQA Model Performance via Correlation Surface

arXiv:2601.21738v1 Announce Type: new Abstract: Evaluation of Image Quality Assessment (IQA) models has long been dominated by global correlation metrics, such as Pearson Linear Correlation...

Software Psychology

arXiv CS Jan 30

Why Adam Works Better with $\beta_1 = \beta_2$: The Missing Gradient Scale Invariance Principle

arXiv:2601.21739v1 Announce Type: new Abstract: Adam has been at the core of large-scale training for almost a decade, yet a simple empirical fact remains unaccounted...

Engineering Psychology