Knowledge catalogue
arXiv:2609.33628v1 Announce Type: new Abstract: Prompt injection is a leading security risk for LLMs and LLM-based applications such as agents. State-of-the-art red-teaming methods for prompt injectio
arXiv:2609.34439v1 Announce Type: new Abstract: Medical vision-language pre-training largely follows a visit-level image-report matching paradigm, aligning paired images and reports at individual visi
arXiv:2609.32962v1 Announce Type: new Abstract: Gradient clipping is widely used to stabilize training, but it need not improve the statistical accuracy of averaged SGD, even under heavy-tailed noise.
arXiv:2609.35476v1 Announce Type: new Abstract: Embodied co-painting requires a robot to repeatedly update a shared physical canvas while human intent evolves over interaction. Existing reference-driv
arXiv:2607.17568v2 Announce Type: replace Abstract: Resource-constrained deployment requires sustaining large language model (LLM) capabilities as models scale under fixed memory and computation budge
arXiv:2609.33807v1 Announce Type: new Abstract: How well can general-purpose multimodal models turn visual understanding and reasoning into embodied manipulation via executable code? We introduce Code
arXiv:2606.18286v2 Announce Type: replace Abstract: Supervised fine-tuning of code LLMs typically applies uniform cross-entropy loss to all response tokens, implicitly assuming that every token provid
Codex Security Cloud is getting a major upgrade, with access to cyber-capable models through Daybreak Blue included by default. It scans entire GitHub repos, continuously reviews new commits, investig
arXiv:2609.34422v1 Announce Type: new Abstract: Language-model agents increasingly tackle long-horizon tasks whose interaction histories exceed the model's active context. Recent work has begun to use
arXiv:2609.34749v1 Announce Type: new Abstract: Real-world driving is inherently multi-agent, yet most existing driving world models generate observations from a single ego vehicle. Independently exte
arXiv:2609.33671v1 Announce Type: cross Abstract: When must a foundation-model safety gateway generate tokens, and when should it directly output a calibrated decision? We study calibrated standalone
arXiv:2609.34782v1 Announce Type: cross Abstract: Many physical tasks in human environments require collaboration, from assisting a partner to jointly manipulating an object. Yet, existing humanoid be
arXiv:2503.19225v4 Announce Type: replace Abstract: We introduce CoinFT, a capacitive 6-axis force/torque (F/T) sensor that is compact, light, low-cost, and robust with an average root-mean-squared er
arXiv:2609.35315v1 Announce Type: new Abstract: Large Language Model (LLM)-based agents promise to automate scientific discovery, yet exploring the vast hypothesis space remains costly. Existing princ
arXiv:2609.33754v1 Announce Type: new Abstract: Organizations seek analytical value from AI, yet relevant data are often fragmented across organizations and constrained by privacy. This is acute in fi
arXiv:2609.32783v1 Announce Type: cross Abstract: Before a team of robots moves, each proposed step must be checked for collisions with other robots and with obstacles. We present CollisionGAT, a grap
arXiv:2609.35619v1 Announce Type: new Abstract: Incorporating dense visual information into motion planning remains challenging, as geometric planners rely on abstracted scene representations that dis
arXiv:2609.34899v1 Announce Type: cross Abstract: Multi-vector retrievers built on vision-language models lead visual document retrieval (VDR), but they run a multi-billion-parameter query encoder on
arXiv:2609.32935v1 Announce Type: new Abstract: Multi-human multi-robot (MH-MR) teams combine robotic autonomy with human expertise, but effective task allocation requires coordinating scarce, dynamic
arXiv:2609.31989v1 Announce Type: cross Abstract: We study communication between two frozen large language models from different providers, with different tokenizers, accessed through their API endpoi
arXiv:2609.34348v1 Announce Type: new Abstract: Gradient updates on different data generally do not commute: training a language model on two data sources in opposite orders gives different weights, e
arXiv:2609.32228v1 Announce Type: new Abstract: In self-play, a proposer generates verifiable tasks to train a solver. Proposer rewards often depend on the solver's success rate, but equally difficult
arXiv:2609.33310v1 Announce Type: new Abstract: Whole-body compliant control is essential for deploying heavy humanoids under high payload in human-centric environments. Most prior force-aware learnin
arXiv:2609.34992v1 Announce Type: new Abstract: Decoding for large language models is typically treated as a collection of isolated sampling strategies, with limited theoretical understanding of the b
arXiv:2609.32566v1 Announce Type: new Abstract: Intelligence is defined in many ways. One of these definitions defines intelligence as the pursuit of learnable novelty. However, learnable novelty can
arXiv:2609.33580v1 Announce Type: new Abstract: Learning-augmented algorithms for metrical task systems (MTS) can exploit predictions of canonical dual values, but existing formulations typically requ
arXiv:2609.34585v1 Announce Type: new Abstract: We propose Tiny Recursive Models for Combinatorial Optimization (ours{}), a general neural method for combinatorial optimization that scales both depth
arXiv:2609.34494v1 Announce Type: new Abstract: Image-to-CAD generation seeks executable parametric programs that recover both the geometry and design intent of a reference object. Existing systems ar
arXiv:2603.08309v3 Announce Type: replace Abstract: Vision Transformers (ViTs) often degrade under distribution shifts because they rely on spurious correlations, such as background cues, rather than
arXiv:2609.33445v1 Announce Type: new Abstract: Concept erasure aims to suppress undesirable knowledge in text-to-image generative models. However, existing robustness evaluations typically rely on re
arXiv:2609.32671v1 Announce Type: cross Abstract: Cross-modal retrieval has been advanced by vision-language pre-trained models that encode images and texts into a shared dense embedding space. While
arXiv:2607.18574v2 Announce Type: replace Abstract: Direct feedback alignment (DFA) trains hidden layers with fixed random projections of the output error. Even when this feedback provides useful cred
arXiv:2609.31792v1 Announce Type: cross Abstract: While Vision-Language-Action (VLA) models perform strongly on manipulation tasks, their responses to invalid task premises remain underexplored. Exist
arXiv:2609.34887v1 Announce Type: cross Abstract: Tabular foundation models (TFMs) provide predictive distributions for regression, but their prediction regions can exhibit undercoverage or overcovera
arXiv:2604.18441v2 Announce Type: replace-cross Abstract: We introduce a robust conformal prediction method based on the half-mass radius P(x):=inf{r>0:P(B(x,r))>1/2}. Its empirical version is exactly
Congrats to @hardmaru @risi1979 @yujin_tang and Risto! Neuroevolution is where I got started in AI, and neuroevolution is still a playground for original ideas with neural networks. This book is worth
🚨🔥 Congress (@RepJoshG @ValerieFoushee @tedlieu) to OpenAI, Anthropic, Google, Meta, SpaceX on their inadequate reporting of their inadequately-controlled agents: “Your companies have since notified t
arXiv:2609.35193v1 Announce Type: new Abstract: Understanding how two entities are connected often requires tracing multi-hop relations across documents to identify intermediate entities and supportin
As query traffic grows, so does the compute bill for embedding it. On a low-power device, a large model may not fit in memory. A smaller query model could reduce that cost, but switching usually means
arXiv:2609.33735v1 Announce Type: new Abstract: Text-guided image editing aims to perform a desired edit while preserving source content unrelated to it. Pretrained rectified-flow models enable traini
arXiv:2609.32952v1 Announce Type: new Abstract: Online safe reinforcement learning (RL) seeks policies that maximize reward while satisfying safety constraints. Reward and safety can induce multimodal
arXiv:2609.34400v1 Announce Type: new Abstract: Safe robot-assisted ultrasound imaging requires a reliable controller able to detect and localize probe--tissue interaction. In this paper, we present a
arXiv:2609.32185v1 Announce Type: new Abstract: Pathology vision-language models (VLMs) are conventionally evaluated by accuracy, but accuracy alone does not measure evidence use: it may conflate data
arXiv:2609.34847v1 Announce Type: new Abstract: Time series prediction is a common application of reservoir computing. When the training and testing time series data contains multiple dynamical regime
arXiv:2601.11016v3 Announce Type: replace-cross Abstract: We propose a framework for contextual distributionally robust optimization (DRO) that considers the causal and continuous structure of the und
arXiv:2609.32328v1 Announce Type: new Abstract: Data unlearning in diffusion models aims to remove the influence of specific training examples without suppressing the broader concepts they represent.
arXiv:2609.32771v1 Announce Type: new Abstract: Adapting pretrained models to new data can cause catastrophic forgetting of previously learned behavior. When only a few past samples remain, they give
arXiv:2609.32026v1 Announce Type: cross Abstract: Physical systems evolve continuously in time, yet their states are typically observed only at discrete times. Generating trajectories consistent with
arXiv:2609.35083v1 Announce Type: cross Abstract: Biological machine learning was long bottlenecked by the ability to synthesize designed DNA. Variational synthesis models control chemical reactions t
arXiv:2604.01601v2 Announce Type: replace Abstract: In-context fine-tuning (IC-Train), training an LLM with labeled examples in-context, is increasingly used in place of standard fine-tuning for domai
arXiv:2609.35603v1 Announce Type: new Abstract: Joint-embedding predictive architectures enable planning with latent world models, but accurate transition prediction alone does not ensure that the pla
arXiv:2609.32038v1 Announce Type: new Abstract: Extended Reality (XR) users do not directly perceive the output of a rendering engine. Instead, rendered images pass through a post-processing pipeline
arXiv:2609.32436v1 Announce Type: cross Abstract: Mechanisms for generating GNN explanations are crucial for building trust and mitigating biases in Graph Neural Networks (GNNs), especially in high-st
arXiv:2609.33810v1 Announce Type: cross Abstract: Autoregressive text-to-speech (TTS) systems synthesize natural speech but, once trained, offer little control over speaking rate. We show that speakin
arXiv:2609.34807v1 Announce Type: new Abstract: Spatially conditioned diffusion models can embed words and contours in natural-looking images, but vision-language models (VLMs) may fail to recognize t
arXiv:2609.34196v1 Announce Type: new Abstract: Modern vision-language models (VLMs) achieve strong performance across a broad range of multimodal tasks, yet still struggle with visual questions that
arXiv:2609.31786v1 Announce Type: cross Abstract: Maximum likelihood estimation of Markov deterioration hazard models for infrastructure asset management is sensitive to the numerical conditioning of
arXiv:2609.33152v1 Announce Type: new Abstract: Muon is emerging as a promising alternative to AdamW for large-scale neural network training, yet theoretical understanding of its practical implementat
arXiv:2609.32393v1 Announce Type: cross Abstract: Plug-and-Play (PnP) methods solve imaging inverse problems by incorporating deep denoisers into iterative optimization algorithms. Although practical
arXiv:2609.35418v1 Announce Type: cross Abstract: This paper studies convex optimization when the gradient cannot be evaluated exactly, but only approximated by a hierarchy of algorithms whose compute