Search: “reasoning”

arXiv:2605.12500v1 Announce Type: new Abstract: Recent large vision-language models (VLMs) remain fundamentally constrained by a persistent dichotomy: understanding and generation are treated as disti

arXiv:2605.11680v1 Announce Type: new Abstract: We introduce ShapeCodeBench, a synthetic benchmark for perception-to-program reconstruction: given a rendered raster image, a model must emit an executa

arXiv:2603.15759v2 Announce Type: replace-cross Abstract: Robot learning requires adaptation methods that improve reliably from limited, mixed-quality interaction data. This is especially challenging

arXiv:2605.12456v1 Announce Type: cross Abstract: We introduce TextSeal, a state-of-the-art watermark for large language models. Building on Gumbel-max sampling, TextSeal introduces dual-key generatio

arXiv:2605.11167v1 Announce Type: new Abstract: Existing multi-model and tool-augmented systems communicate by generating text, serializing every exchange through the output vocabulary. Can two pretra

arXiv:2507.13841v2 Announce Type: replace Abstract: Good storytelling involves surprise -- unpredictability in how the story unfolds -- and sense-making, the requirement that the story forms a coheren

In the healthcare and life sciences industries, speed saves lives, but meeting regulatory requirements and other administrative burdens often pumps the brakes for manufacturers of software as a medica

arXiv:2605.11974v1 Announce Type: new Abstract: Large Language Models (LLMs) suffer from order bias, where their performance is affected by the arrangement order of input elements. This unfairness lim

arXiv:2510.03853v4 Announce Type: replace Abstract: We present UGround, a extbf{U}nified visual extbf{Ground}ing paradigm that dynamically selects intermediate layers across extbf{U}nrolled transforme

arXiv:2605.11170v1 Announce Type: new Abstract: Noise-based certified machine unlearning currently faces a hard ceiling: the noise magnitude required to certify unlearning typically destroys model uti

arXiv:2605.11334v1 Announce Type: cross Abstract: LLM-as-Judge systems are widely deployed for automated evaluation, yet practitioners lack reliable methods to know when a judge's verdict should be tr

🎉 We published a new AI safety study: shopping agents fall for whimsical attacks and lose money. A whimsical attack is an absurd scenario a human would never try on another human. In one run, GPT-5.1

We’re training models wrong and it’s due to chatGPT. Even the modern coding agents used daily still use message-based exchanges: They send messages to users, to themselves (CoT) and to tools, and rece

arXiv:2605.12021v1 Announce Type: new Abstract: Many image understanding tasks involve identifying what is present and where it appears. However, tasks that address where, such as object discovery, de

arXiv:2605.12162v1 Announce Type: new Abstract: Effectively handling the interplay between spatial perception and action generation remains a critical bottleneck in robotic manipulation. Existing meth

arXiv:2605.08504v1 Announce Type: new Abstract: We investigate the origins of massive activations in large language models (LLMs) and identify a specific layer named the extbf{Massive Emergence Layer

arXiv:2605.08530v1 Announce Type: new Abstract: Millimeter-wave (mmWave) radar has emerged as a promising sensing modality for human perception due to its robustness under challenging environmental co

arXiv:2605.08647v1 Announce Type: cross Abstract: Multi-agent systems achieve state-of-the-art outcomes through peer collaboration. However, when an agent in the pipeline silently drops a constraint,

Agentic security operations startup Exaforce Inc. today announced that it has raised 125 million in new funding to scale up its AI-driven security operations platform and expand its real-time threat r

AI has a secondary intent problem. I was negotiating a rent renewal (I live in NYC and they raised rent 10% because I guess they were bored). The property manager said she would 'do everything she cou

arXiv:2604.08178v2 Announce Type: replace Abstract: In classical Reinforcement Learning from Human Feedback (RLHF), Reward Models (RMs) serve as the fundamental signal provider for model alignment. As

arXiv:2605.10415v1 Announce Type: new Abstract: Large language models for subjectivity analysis are typically trained with aggregated labels, which compress variations in human judgment into a single

arXiv:2605.09150v1 Announce Type: new Abstract: Poker is an imperfect information game that has served as a long-standing benchmark for decision-making under uncertainty. To maximize utility beyond th

arXiv:2605.10584v1 Announce Type: cross Abstract: With the detection of gravitational waves (GWs), multi-messenger astronomy has opened a new window for advancing our understanding of astrophysics, de

arXiv:2605.10339v1 Announce Type: new Abstract: The advancement of Large Language Models (LLMs) has enabled their application in personalized dialogue systems. We present an extended annotation scheme

arXiv:2605.10397v1 Announce Type: cross Abstract: Visual anomaly detection (VAD) is crucial in many real-world fields, such as industrial inspection, medical imaging, infrastructure monitoring, and re

arXiv:2605.09492v1 Announce Type: cross Abstract: Large language models (LLMs) often suffer from hallucinations due to error accumulation in autoregressive decoding, where suboptimal early token choic

arXiv:2605.10153v1 Announce Type: cross Abstract: Explainable AI (XAI) has achieved remarkable success in image classification, yet the audio domain lacks equally mature solutions. Current methods app

arXiv:2603.10126v2 Announce Type: replace-cross Abstract: We propose a standalone autoregressive (AR) Action Expert that generates actions as a continuous causal sequence while conditioning on refresh

arXiv:2605.08831v1 Announce Type: new Abstract: In flexible assembly systems, existing task planning methods require a time-consuming configuration process by multiple experts to establish a productio

arXiv:2509.08031v3 Announce Type: replace-cross Abstract: Large Audio Language Models (LALMs) are rapidly advancing, but evaluating them remains challenging due to inefficient and non-standardized too

arXiv:2604.02151v2 Announce Type: replace Abstract: We consider multi-objective reinforcement learning problems where objectives come from an identical family -- such as the class of reachability obje

arXiv:2605.08110v1 Announce Type: cross Abstract: Low-Rank Adaptation (LoRA) has become the standard for fine-tuning large pre-trained models at reduced computational cost. However, its low-rank point

arXiv:2605.10865v1 Announce Type: new Abstract: Industrial Computer-Aided Design (CAD) code generation requires models to produce executable parametric programs from visual or textual inputs. Beyond r

arXiv:2605.09285v1 Announce Type: new Abstract: Null-space-based methods have garnered considerable attention in model editing by constraining updates to the null space of the pre-existing knowledge r

arXiv:2605.10034v1 Announce Type: new Abstract: Recent Autonomous Driving (AD) works such as GigaFlow and PufferDrive have unlocked Reinforcement Learning (RL) at scale as a training strategy for driv

arXiv:2605.10438v1 Announce Type: cross Abstract: Current 3D tokenizers largely treat representation as spatial compression: compact codes reconstruct surface geometry, but leave component ownership a

arXiv:2605.09134v1 Announce Type: new Abstract: Reinforcement learning for program repair is hindered by sparse execution feedback and coarse sequence-level rewards that obscure which edits actually f

arXiv:2512.18880v2 Announce Type: replace-cross Abstract: Accurate estimation of item (question or task) difficulty is critical for educational assessment but suffers from the cold start problem. Whil

arXiv:2512.04949v3 Announce Type: replace-cross Abstract: Agents capable of accomplishing complex tasks through multiple interactions with the environment have emerged as a popular research direction.

arXiv:2605.08173v1 Announce Type: new Abstract: The generalization performance (GP) of deep learning-based arbitrary-scale image super-resolution (ASISR) methods is subject to limited training dataset

arXiv:2605.08383v1 Announce Type: new Abstract: Philosophical accounts of persuasion often assume that shared evidence and rational argumentation should lead to a convergence of views between peers, y

arXiv:2605.10163v1 Announce Type: cross Abstract: Recent work on causal abstraction, in particular graphical approaches focusing on causal structure between clusters of variables, aims to summarize a

arXiv:2605.08399v1 Announce Type: new Abstract: Tool-augmented language models can extend small language models with external executable skills, but scaling the tool library creates a coupled challeng

arXiv:2605.10462v1 Announce Type: new Abstract: Formalisation is the process of writing system requirements in a formal language. These requirements mostly originate in Natural Language. In the field

arXiv:2605.10123v1 Announce Type: new Abstract: Complex-valued Transformers have largely inherited softmax attention from real-valued architectures. However, row-normalised token competition is not ne