Knowledge catalogue
arXiv:2609.32060v1 Announce Type: new Abstract: How many bits does a network need before its accuracy collapses, and how does this grow with depth? We study the precision floor, the perturbation level
arXiv:2609.32534v1 Announce Type: new Abstract: Depth is a natural way to increase the computational capacity in Transformers, yet the contribution of deeper layers can diminish as depth grows larger.
arXiv:2609.32797v1 Announce Type: new Abstract: Derivative-informed training improves neural operators by directly supervising their input-output sensitivities, which is crucial when neural operators
arXiv:2509.01887v2 Announce Type: replace-cross Abstract: We study the design of interventions for causal discovery in simple structural causal models whose causal graphs are directed mixed graphs (DM
arXiv:2609.35022v1 Announce Type: new Abstract: The integration of deep learning models into image preprocessing pipelines such as super-resolution introduces a largely unexplored attack vector for ad
arXiv:2609.31681v1 Announce Type: new Abstract: Devnagari is a one of the ancient language of the Indian subcontinent consisting of 36 vowels, 14 consonants and 10 numerals. The accurate recognition o
arXiv:2609.33048v1 Announce Type: new Abstract: Understanding how individual brain development unfolds over childhood requires modeling developmental trajectories from sparse longitudinal observations
arXiv:2609.32193v1 Announce Type: new Abstract: Vision Language Action (VLA) models condition actions directly on current visual and language context, without an explicit account of how the scene evol
arXiv:2609.35318v1 Announce Type: new Abstract: Human videos offer a scalable source of demonstrations for dexterous robot manipulation. However, existing human-to-simulation-to-robot (Human2Sim2Robot
arXiv:2609.35761v1 Announce Type: new Abstract: Mobile bimanual dexterous manipulation requires continuous coordination of locomotion, whole-body motion, and finger-level dexterity within a single tra
arXiv:2609.33882v1 Announce Type: new Abstract: Glove-based motion capture is emerging as a scalable approach to collecting dexterous-hand demonstration data. However, due to the kinematic gap between
arXiv:2609.34286v1 Announce Type: new Abstract: World models for manipulation are typically trained from video, yet the events that determine how manipulation unfolds, such as making and releasing con
arXiv:2609.34724v1 Announce Type: new Abstract: Learning dexterous humanoid loco-manipulation from human demonstrations requires transferring not only human motion, but also the coordinated interactio
arXiv:2609.35096v1 Announce Type: new Abstract: Deepfake detection methods have become increasingly effective yet most provide limited insight into the evidence behind their predictions. However, in f
arXiv:2609.31885v1 Announce Type: new Abstract: Autonomous micro-mobility vehicles (MMVs) such as wheelchairs, scooters, and bicycles have the potential to improve mobility access and support safe low
arXiv:2606.19939v2 Announce Type: replace Abstract: Handwritten Mathematical Expression Generation (HMEG) is challenging due to the complex two-dimensional layouts and long-range structural dependenci
arXiv:2609.32363v1 Announce Type: new Abstract: Probabilistic time series forecasting requires modeling and predicting complex and time-varying distributions. Recently, Denoising Diffusion Probabilist
arXiv:2609.33930v1 Announce Type: new Abstract: Extending forecast lead times while maintaining predictive skill remains a major challenge in environmental forecasting. We investigate diffusion-based
arXiv:2508.15641v2 Announce Type: replace Abstract: Precise temporal grounding requires distinguishing when a queried event occurs from when its participating entities are merely visible. We propose D
arXiv:2609.33803v1 Announce Type: new Abstract: Reward models underpin the alignment of large language models, yet the dominant designs reduce each prompt--response pair to a point estimate or to a di
arXiv:2511.15990v2 Announce Type: replace-cross Abstract: Digital agriculture is transforming the way we grow food by utilizing technology to make farming more efficient, sustainable, and productive.
arXiv:2609.31750v1 Announce Type: new Abstract: Recognition of student affective states from classroom video is constrained by a class imbalance problem whose true nature is, we argue, under-analyzed.
arXiv:2609.35444v1 Announce Type: new Abstract: Robust skeleton-based action recognition requires representations that capture a wide spectrum of motions, from subtle to moderate and strong ones. Exis
arXiv:2609.32579v1 Announce Type: new Abstract: A linear projection can reduce the dimension of query and key vectors without updating the pretrained model, but it remains unclear which training objec
arXiv:2510.17611v3 Announce Type: replace Abstract: Unsupervised anomaly detection (UAD) has evolved from specialized single-class models to unified multi-class models, yet existing multi-class models
arXiv:2609.33298v1 Announce Type: new Abstract: In many settings, post-training need not create the target behavior from scratch: the base model can already produce it, but not reliably. This shifts p
arXiv:2609.34279v1 Announce Type: new Abstract: Self-evolving language models improve by generating tasks and learning from their own feedback, but adapting the task generator often requires a separat
arXiv:2609.34651v1 Announce Type: new Abstract: Generating high-quality UV textures for 3D meshes remains challenging. Multi-view projection pipelines suffer from occlusion and view inconsistency, and
arXiv:2511.01813v3 Announce Type: replace-cross Abstract: We introduce disciplined biconvex programming (DBCP), a modeling framework for specifying and solving biconvex optimization problems. Biconvex
arXiv:2609.33527v1 Announce Type: new Abstract: Parameter space symmetries are important for understanding neural networks' loss landscape, training dynamics, and generalization. However, systematical
arXiv:2603.18257v3 Announce Type: replace Abstract: When an RL agent's observations contain distractors driven by the same confounders as its true state, observational data alone cannot identify which
arXiv:2609.31693v1 Announce Type: new Abstract: Invisible image watermarks are commonly evaluated against benign postprocessing operations such as compression, resizing, blur, and color changes. These
arXiv:2609.35240v1 Announce Type: new Abstract: Artificial-intelligence studies using computed tomography (CT) for lung cancer are often broadly labelled 'prediction' despite addressing clinically dis
arXiv:2609.34629v1 Announce Type: new Abstract: Many complex systems are observed only through temporally unpaired distribution snapshots, making trajectory-based dynamical learning difficult without
arXiv:2609.32663v1 Announce Type: new Abstract: The memory usage and decoding latency of LLM inference grow rapidly with context length. To reduce these costs, key-value (KV) cache compression methods
arXiv:2609.35699v1 Announce Type: new Abstract: Distillation attacks copy the reasoning capabilities of closed-source large language models, allowing bad actors to replicate state-of-the-art performan
arXiv:2609.34408v1 Announce Type: new Abstract: Large Vision-Language Models (LVLMs) have shown strong promise for multimodal reasoning, yet often struggle with tasks requiring concepts beyond what is
arXiv:2609.32971v1 Announce Type: new Abstract: Accurate forecasting of river discharge and floods is very challenging. River dynamics are affected by storage, meteorological forcing, and flow propaga
arXiv:2609.34503v1 Announce Type: new Abstract: Parameter-efficient adaptation enables continual learners to acquire task-specific knowledge through compact model updates while maintaining strong with
arXiv:2609.31653v1 Announce Type: cross Abstract: Sentiment analysis is a common tool for converting unstructured text into quantitative signals in finance and risk management. Yet most applications r
arXiv:2609.32727v1 Announce Type: new Abstract: We study distributionally robust reinforcement learning (DR-RL) in the average-reward setting under weak communication. Our main result provides finite-
arXiv:2605.06553v2 Announce Type: replace Abstract: Modern generative models can produce high-quality samples, but independent samples under the same condition often yields highly similar outputs. Par
arXiv:2607.11684v2 Announce Type: replace-cross Abstract: Existing contextual multinomial logit (MNL) bandits model relevance-driven choice but ignore the potential benefits of within-assortment diver
arXiv:2609.34838v1 Announce Type: new Abstract: On-policy distillation (OPD) trains student agents through teacher supervision on their own interactions with an environment. However, in asynchronous m
arXiv:2609.33083v1 Announce Type: new Abstract: Sparse video understanding often requires selecting a small set of visual evidence under a fixed frame budget. Most sparse selectors allocate this budge
arXiv:2609.34742v1 Announce Type: new Abstract: Recent studies suggest that large language models encode emotion concepts as structured internal representations, but most existing work focuses on text
arXiv:2602.00449v2 Announce Type: replace-cross Abstract: Latent Chain-of-Thought aims to enable step-by-step computation without emitting long rationales, yet its internal mechanisms remain unclear.
arXiv:2609.35039v1 Announce Type: new Abstract: Vision-Language-Action (VLA) policies trained with behavior cloning or flow matching are optimized to output an action trajectory, but they cannot expre
arXiv:2609.35364v1 Announce Type: new Abstract: Many temporal link predictors summarize past interactions through learned node representations. We examine whether simple counts of recurring interactio
arXiv:2609.33791v1 Announce Type: new Abstract: Since the advent of knowledge distillation, KL divergence has been the standard loss in distillation. Recently, on-policy distillation (OPD) has emerged
arXiv:2609.34058v1 Announce Type: new Abstract: AI systems often feel brittle and fragmented. A large language model (LLM) may correctly explain a concept but fail to apply it, or follow safety instru
kinda realized i type things into AI chats that i'd never say to a friend. money problems, work stuff, health worries, random late-night thoughts lol then i found out all of it is stored on their serv
arXiv:2609.33707v1 Announce Type: cross Abstract: Vision-language-action (VLA) models adapt pretrained vision-language models (VLMs) for closed-loop robot control, transferring their perceptual and se
arXiv:2609.33536v1 Announce Type: new Abstract: Large language model (LLM) serving is increasingly constrained by the GPU memory consumed by key-value (KV) caches. Existing compression, eviction, and
arXiv:2609.31680v1 Announce Type: new Abstract: Joint-embedding predictive architectures (JEPA) have emerged as a promising self-supervised pretraining paradigm for time series, learning representatio
arXiv:2609.34621v1 Announce Type: new Abstract: Native 3D texture generation synthesizes colors directly in 3D space for a given geometry, conditioned on multi-view reference images. It is generally b
We did a study: does post training actually make LLMs funnier? We used open models that publish every stage of post training, so we could compare a base model with the future models it became: Tulu 3
arXiv:2609.32525v1 Announce Type: new Abstract: The widely used Transolver family of neural operators is based on physics-attention, which softly assigns the points of an unstructured mesh to a small
arXiv:2609.31985v1 Announce Type: new Abstract: Vision-language pretraining objectives differ in the spatial granularity of their supervision, yet the implications of this distribution for chest X-ray
arXiv:2609.31882v1 Announce Type: new Abstract: Reward-based diffusion fine-tuning faces practical challenges when desirable outcomes are rare or conditioning corrections are costly to estimate. In th