Self-Distilled Agentic Reinforcement Learning
arXiv:2605.15155v1 Announce Type: cross Abstract: Reinforcement learning (RL) has emerged as a central paradigm for post-training LLM agents, yet its trajectory-level reward signal provides only coars
Knowledge catalogue
arXiv:2605.15155v1 Announce Type: cross Abstract: Reinforcement learning (RL) has emerged as a central paradigm for post-training LLM agents, yet its trajectory-level reward signal provides only coars
ArXiv will ban authors for one year if they submit papers containing obviously AI-generated content , with examples including hallucinated citations, placeholder text, or chatbot meta-comments left in
arXiv:2605.14205v1 Announce Type: new Abstract: LLM-based web agents can navigate live storefronts, yet they often collapse to a single 'average buyer' policy, failing to capture the heterogeneous and
arXiv:2605.14089v1 Announce Type: new Abstract: In recent years, a variety of powerful LLM-based agentic systems have been applied to automate complex tasks through task orchestration. However, existi
arXiv:2605.14937v1 Announce Type: cross Abstract: Predictive world models enable agents to model scene dynamics and reason about the consequences of their actions. Inspired by human perception, object
arXiv:2605.14284v1 Announce Type: new Abstract: Comparative evaluation of multiple dynamic treatment policies is essential for healthcare and policy decisions, yet conventional longitudinal causal inf
arXiv:2605.15074v1 Announce Type: new Abstract: Reliable pose estimation in previously unseen environments is a fundamental capability of autonomous systems. Existing LiDAR odometry methods typically
arXiv:2605.15044v1 Announce Type: cross Abstract: As audio-first agents become increasingly common in physical AI, conversational robots, and screenless wearables, audio large language models (audio-L
arXiv:2605.14566v1 Announce Type: new Abstract: Medical image segmentation remains challenging in low-data regimes, where scarce annotations often yield poor generalization and ambiguous boundaries wi
Gary Marcus criticizes Geoffrey Hinton for allegedly misrepresenting or fabricating a quote attributed to him in a debate about AI. The post sarcastically compares the frequency of strawmanning argume
arXiv:2512.09115v2 Announce Type: replace Abstract: High-resolution imagery is often hindered by limitations in sensor technology, atmospheric conditions, and costs. Such challenges occur in satellite
arXiv:2605.14440v1 Announce Type: new Abstract: Partially Observable Markov Decision Processes (POMDPs) are the standard framework for decision-making under uncertainty. While sampling-based methods s
arXiv:2605.14396v1 Announce Type: new Abstract: Autonomous vehicles depend on online HD map construction to perceive lane boundaries, dividers, and pedestrian crossings -- safety-critical road element
arXiv:2605.14879v1 Announce Type: cross Abstract: A plethora real-world environments require agents to compete repeatedly for the same limited resource, calling for a temporal notion of fairness judge
arXiv:2605.13909v1 Announce Type: cross Abstract: Negotiation is a central mechanism of economic exchange, shaping markets, procurement, labor agreements, and resource allocation. It is also a canonic
arXiv:2605.14418v1 Announce Type: cross Abstract: 'Oh-Oh, yes, I'm the great pretender. Pretending that I'm doing well. My need is such, I pretend too much...' summarizes the state in the area of jail
arXiv:2510.00231v2 Announce Type: replace-cross Abstract: KV cache compression promises increased throughput and efficiency with negligible loss in performance. While the gains in throughput are indis
arXiv:2605.14705v1 Announce Type: new Abstract: Sign language is the primary language for many Deaf and Hard-of-Hearing (DHH) signers, yet most conversational AI systems still mediate interaction thro
arXiv:2605.14276v1 Announce Type: cross Abstract: Diffusion models generate samples by denoising along the score of a perturbed target distribution. In practice, one trains a neural diffusion model, w
arXiv:2605.15134v1 Announce Type: new Abstract: Estimating how often an ML model will fail at deployment scale is central to pre-deployment safety assessment, but a feasible evaluation set is rarely l
arXiv:2605.14731v1 Announce Type: cross Abstract: Speech-driven gestures and facial animations are fundamental to expressive digital avatars in games, virtual production, and interactive media. Howeve
arXiv:2605.14643v1 Announce Type: new Abstract: Deep learning methods based on backward stochastic differential equations (BSDEs) have emerged as competitive alternatives to physics-informed neural ne
arXiv:2605.14626v1 Announce Type: new Abstract: RGB-T semantic segmentation requires strictly aligned VIS-IR-Label triplets; however, such aligned triplet data are often scarce in real-world scenarios
arXiv:2603.11042v2 Announce Type: replace-cross Abstract: Generating music that temporally aligns with video events is challenging for existing text-to-music models, which lack fine-grained temporal c
Very thoughtful post on the role and success of open source, including potential applications for AVs and AI. Long, but worth the read. -- 'Open source is no longer just how good software gets built.
arXiv:2605.15195v1 Announce Type: new Abstract: Recent feed-forward reconstruction models, such as VGGT, have proven competitive with traditional optimization-based reconstructors while also providing
arXiv:2602.02994v2 Announce Type: replace Abstract: Reinforcement learning has emerged as a principled post-training paradigm for Temporal Video Grounding (TVG) due to its on-policy optimization, yet
arXiv:2605.14710v1 Announce Type: cross Abstract: Deep learning and multi-modal fusion have demonstrated transformative potential in medical diagnosis by integrating diverse data sources. However, acc
arXiv:2510.11282v2 Announce Type: replace Abstract: Accurate spatiotemporal traffic forecasting is a critical prerequisite for proactive resource management in dense urban mobile networks. While large
arXiv:2605.14972v1 Announce Type: cross Abstract: A fundamental limitation of Text-to-Code is that no guarantee can be obtained about the correctness of the generated code. Therefore, to ensure its co
arXiv:2605.15182v1 Announce Type: new Abstract: Camera-controlled video generation has made substantial progress, enabling generated videos to follow prescribed viewpoint trajectories. However, existi
“we are below rock bottom now”. Trump admin continues to piss away America’s scientific lead for no good reason. @elonmusk stands by, does nothing. NEW: A serious staffing shortage — of the Trump admi
arXiv:2511.02776v2 Announce Type: replace Abstract: Recent progress in large-scale robotic datasets and vision-language models (VLMs) has advanced research on vision-language-action (VLA) models. Howe
Gary Marcus highlighted a concerning problem in 2025 where approximately 140,000 fabricated citations were identified, likely referring to false or hallucinated references generated by AI systems in a
arXiv:2605.12608v1 Announce Type: new Abstract: Object detection in adverse weather is critical for the safety of autonomous vehicles; however, the scarcity of labelled, real-world foggy data remains
arXiv:2605.12719v1 Announce Type: cross Abstract: The continual assurance of safety and performance of automated driving systems (ADSs) poses significant challenges. ADSs operate in complex, dynamic,
arXiv:2605.13161v1 Announce Type: new Abstract: Efficient transfer learning methods for large-scale vision-language models (e.g., CLIP) enable strong few-shot transfer, yet existing adaptation methods
arXiv:2605.13639v1 Announce Type: new Abstract: In this paper, we establish last-iterate convergence rates for off-policy actor--critic methods in reinforcement learning. In particular, under a single
arXiv:2605.12569v1 Announce Type: cross Abstract: Global navigation satellite system (GNSS) interference poses a serious threat to reliable positioning, especially in indoor and multipath-rich environ
arXiv:2605.12878v1 Announce Type: cross Abstract: We propose Adam-SHANG, a Lyapunov-guided Adam-type method that couples momentum, adaptive preconditioning, and a curvature-aware correction through a
arXiv:2605.12917v1 Announce Type: new Abstract: Deep learning models for medical imaging often exhibit overconfidence, creating safety risks in ambiguous diagnostic scenarios. While Conformal Predicti
arXiv:2605.13702v1 Announce Type: new Abstract: Strategic mine production scheduling under geological uncertainty is conventionally formulated as a stochastic optimization problem in which a fixed ext
arXiv:2605.12771v1 Announce Type: cross Abstract: Multi-objective reinforcement learning in robotic domains requires balancing complex, non-convex trade-offs between conflicting objectives. While line
arXiv:2602.01629v2 Announce Type: replace Abstract: Rigorous uncertainty quantification is essential for the safe deployment of autonomous systems in unconstrained environments. Conformal Prediction (
arXiv:2501.10598v3 Announce Type: replace Abstract: We study the problem of learning optimal policies in finite-horizon Markov Decision Processes (MDPs) using low-rank reinforcement learning (RL) meth
arXiv:2605.12532v1 Announce Type: cross Abstract: Conventional algorithmic trading systems are grounded in deterministic heuristics or offline-trained statistical models that cannot adapt to the seman
arXiv:2408.12935v4 Announce Type: replace Abstract: AI Safety is an emerging area of critical importance to the safe adoption and deployment of AI systems. With the rapid proliferation of AI and espec
arXiv:2602.02977v2 Announce Type: replace-cross Abstract: Vision-language models such as CLIP often struggle to faithfully understand long, detail-rich captions, relying on dominant scene cues while o
arXiv:2605.13744v1 Announce Type: new Abstract: Image restoration is an inherently ill posed inverse problem. Equivariant networks that embed geometric symmetry priors can mitigate this ill posedness
arXiv:2605.13724v1 Announce Type: cross Abstract: Few-step video generation has been significantly advanced by consistency distillation. However, the performance of consistency-distilled models often
arXiv:2512.10931v3 Announce Type: replace Abstract: Many state-of-the-art LLMs are trained to think before giving their answer. Reasoning can greatly improve language model capabilities, but it also m
arXiv:2602.02560v2 Announce Type: replace-cross Abstract: Lung cancer remains the leading cause of cancer mortality, driving the development of automated screening tools to alleviate radiologist workl
arXiv:2605.06390v2 Announce Type: replace Abstract: A leading proposal for aligning artificial superintelligence (ASI) is to use AI agents to automate an increasing fraction of alignment research as c
arXiv:2601.15161v2 Announce Type: replace-cross Abstract: Large Language Models (LLMs) are increasingly used for clinical decision support, where hallucinations and unsafe suggestions may pose direct
arXiv:2605.12726v1 Announce Type: new Abstract: Final-token safety probes monitor a single hidden state after prompt prefill, but jailbreak prompts can contain probe-visible unsafe evidence distribute
arXiv:2605.12730v1 Announce Type: new Abstract: Existing AI systems for modeling human behavior operate at the level of individuals or detect events after they occur. As a result, they systematically
arXiv:2605.12710v1 Announce Type: new Abstract: Residual risk metrics have recently been introduced to assess the safety implications of automated driving systems. Existing approaches typically assume
arXiv:2605.13574v1 Announce Type: cross Abstract: This study investigates deep self-disclosure toward generative AI by examining perceived non-humanity and structural similarity as psychological facto
arXiv:2605.13525v1 Announce Type: cross Abstract: Automated driving has made remarkable progress, yet situations still arise where human intervention is necessary. Teleoperation provides a scalable so
arXiv:2605.13794v1 Announce Type: cross Abstract: We present BlitzGS, a distributed 3DGS framework that reduces active Gaussian workload for fast city-scale reconstruction. BlitzGS manages this worklo