Knowledge catalogue
'In the past nine months, the United States has produced more AI legislation than in the prior decade,' write @JeffSonnenfeld, @GaryMarcus, and Stephen Henriques in a commentary piece for Fortune. 'No
arXiv:2605.14967v1 Announce Type: new Abstract: Supervised fine-tuning (SFT) provides the standard approach for teaching LLMs new behaviors from offline expert demonstrations. However, standard SFT un
arXiv:2605.14062v1 Announce Type: new Abstract: While synthetic data generation with large language models (LLMs) is widely used in post-training pipelines, existing approaches typically generate full
arXiv:2605.14278v1 Announce Type: new Abstract: Aligning streaming autoregressive (AR) video generators with human preferences is challenging. Existing reinforcement learning methods predominantly rel
arXiv:2605.14531v1 Announce Type: new Abstract: This work reformulates language generation as a stochastic optimal control problem, providing a unified theoretical perspective to analyze autoregressiv
arXiv:2605.15054v1 Announce Type: new Abstract: Vision-language models (VLMs) have recently emerged as a promising paradigm for video anomaly detection (VAD) due to their strong visual reasoning abili
arXiv:2605.14539v1 Announce Type: new Abstract: Reinforcement Learning with Verifiable Rewards (RLVR) has emerged as an effective paradigm for improving the reasoning capabilities of large language mo
arXiv:2605.15113v1 Announce Type: new Abstract: Reinforcement learning from verifiable rewards (RLVR) suffers from sparse outcome signals, creating severe exploration bottlenecks on complex reasoning
arXiv:2605.14392v1 Announce Type: new Abstract: We pursue a vision for self-improving language models in which the model does not merely generate problems or traces to imitate, but constructs the envi
arXiv:2605.14571v1 Announce Type: cross Abstract: Observing touch on another's body can elicit corresponding tactile sensations in the observer, a phenomenon termed mirror touch that supports empathy
arXiv:2605.15108v1 Announce Type: cross Abstract: Off-policy evaluation (OPE) estimates the value of a target treatment policy (e.g., a recommender system) using data collected by a different logging
arXiv:2605.14874v1 Announce Type: new Abstract: Virtual Try-On (VTON) aims to synthesize photorealistic images of garments precisely aligned with a person's body and pose. Current diffusion-based meth
arXiv:2605.14201v1 Announce Type: cross Abstract: Vision-language-action (VLA) models are effective as end-to-end motion planners, but can be brittle when evaluated in closed-loop settings due to bein
arXiv:2605.14087v1 Announce Type: new Abstract: Large Language Models (LLMs), when trained on web-scale corpora, inherently absorb toxic patterns from their training data. This leads to ``toxic degene
arXiv:2605.14744v1 Announce Type: cross Abstract: Large language models in regulated financial workflows are governed by natural-language policies that the same model interprets, creating a principal-
arXiv:2511.02271v2 Announce Type: replace Abstract: Medical Report Generation (MRG) is a key part of modern medical diagnostics, as it automatically generates reports from radiological images to reduc
arXiv:2605.15172v1 Announce Type: cross Abstract: Backdoor attacks pose a serious security threat to large language models (LLMs), which are increasingly deployed as general-purpose assistants in safe
arXiv:2605.14530v1 Announce Type: new Abstract: Large diffusion vision-language models (LDVLMs) have recently emerged as a promising alternative to autoregressive models, enabling parallel decoding fo
arXiv:2605.14364v1 Announce Type: new Abstract: Continual learning requires models to adapt to new data while preserving previously acquired knowledge. At its core, this challenge can be viewed as pri
arXiv:2605.14550v1 Announce Type: new Abstract: Artificial intelligence in high-stakes tabular domains cannot be evaluated by predictive performance alone, yet current practice still assesses explaina
arXiv:2605.14935v1 Announce Type: new Abstract: We present MSCoT, a multi-scale, coarse-to-fine model for test-time human motion synthesis and control. Unlike recent approaches that rely on multiple i
arXiv:2512.07461v3 Announce Type: replace Abstract: We introduce Native Parallel Reasoner (NPR), a teacher-free framework that enables Large Language Models (LLMs) to self-evolve genuine parallel reas
arXiv:2602.00520v3 Announce Type: replace Abstract: Event stream data often exhibit hierarchical structure in which multiple events co-occur, resulting in a sequence of multisets (i.e., bags of events
arXiv:2605.14252v1 Announce Type: cross Abstract: Spiking neural networks (SNNs), which are brain-inspired and spike-driven, achieve high energy efficiency. However, a performance gap between SNNs and
arXiv:2605.15083v1 Announce Type: cross Abstract: The choice of optimiser is important in deep learning, as it strongly influences model efficiency and speed of convergence. However, many commonly use
arXiv:2605.14721v1 Announce Type: new Abstract: Strong equivalence between knowledge bases ensures the possibility of replacing one with the other without affecting reasoning outcomes, in any given co
arXiv:2605.14260v1 Announce Type: cross Abstract: Conformal prediction is often calibrated with a single pooled threshold, but this can hide cross-group heterogeneity in score distributions and distor
arXiv:2512.01766v2 Announce Type: replace Abstract: Last-layer retraining (LLR) methods -- wherein the last layer of a neural network is reinitialized and retrained on a held-out set following ERM tra
arXiv:2605.14605v1 Announce Type: cross Abstract: Model providers increasingly release open weights or allow users to fine-tune foundation models through APIs. Although these models are safety-aligned
arXiv:2603.24422v2 Announce Type: replace-cross Abstract: Generative Retrieval (GR) has emerged as a promising paradigm for modern search systems. Compared to multi-stage cascaded architecture, it off
Transformer: OpenAI's disavowal of a liability shield in Illinois SB 3444 bill and endorsement of a stronger SB 315 suggest it is open to meaningful AI safety legislation — Transformer Weekly: US-Chin
arXiv:2605.14978v1 Announce Type: new Abstract: Speculative decoding accelerates LLM inference by having a lightweight draft model propose speculative windows of candidate tokens for parallel verifica
arXiv:2605.14765v1 Announce Type: cross Abstract: Persian music, with its unique tonalities, modal systems (Dastgah), and rhythmic structures, presents significant challenges for music generation mode
arXiv:2605.13894v1 Announce Type: cross Abstract: In this work, we introduce a Tropical Axial Attention neural reasoning architecture that replaces vanilla softmax dot-product attention with max-plus
arXiv:2605.14297v1 Announce Type: cross Abstract: We study reinforcement learning in hybrid discrete-continuous action spaces, such as settings where the discrete component selects a regime (or index)
arXiv:2605.15164v1 Announce Type: cross Abstract: This position paper argues that behavioural assurance, even when carefully designed, is being asked to carry safety claims it cannot verify. AI govern
arXiv:2605.14294v1 Announce Type: new Abstract: Formal verification of transformers has become increasingly important due to their widespread deployment in safety-critical applications. Compared to cl
arXiv:2504.01571v2 Announce Type: replace-cross Abstract: We use hierarchical procedural rules for the generation of control maps within the stable diffusion framework to produce photo-realistic archi
arXiv:2605.14758v1 Announce Type: new Abstract: History-dependent policies induced by recurrent neural networks (RNNs) rely on latent hidden state dynamics, making verification in partially observable
arXiv:2504.11703v3 Announce Type: replace-cross Abstract: AI agents interact with external environments through tool calls, exposing them to attacks like indirect prompt injection that can trigger una
arXiv:2605.14443v1 Announce Type: new Abstract: The shift toward interacting with frozen, 'black-box' Large Language Models (LLMs) has transformed prompt engineering from a heuristic exercise into a c
arXiv:2602.07441v2 Announce Type: replace-cross Abstract: Offline reinforcement learning (RL), which optimizes policies using a previously collected static dataset, is an important branch of RL. A pop
arXiv:2602.04289v2 Announce Type: replace Abstract: Modern language models are trained almost exclusively on token sequences produced by a fixed tokenizer, an external lossless compressor often over U
arXiv:2605.15000v1 Announce Type: cross Abstract: Premature closure, or committing to a conclusion before sufficient information is available, is a recognized contributor to diagnostic error but remai
arXiv:2605.15185v1 Announce Type: cross Abstract: Generative video models are increasingly studied as implicit world models, yet evaluating whether they produce physically plausible 3D structure and m
arXiv:2511.17367v2 Announce Type: replace Abstract: Computing worst-case robust strategies in pursuit-evasion games (PEGs) is time-consuming, especially when real-world factors like partial observabil
arXiv:2605.14026v1 Announce Type: cross Abstract: For reinforcement learning in data-scarce domains like real-world robotics, intensive data reuse enhances efficiency but induces overfitting. While pr
arXiv:2510.20206v2 Announce Type: replace Abstract: Prompt design plays a crucial role in text-to-video (T2V) generation, yet user-provided prompts are often short, unstructured, and misaligned with t
arXiv:2605.15190v1 Announce Type: new Abstract: Causal autoregressive video diffusion models support real-time streaming generation by extrapolating future chunks from previously generated content. Di
arXiv:2601.03630v2 Announce Type: replace Abstract: This paper presents the first systematic comparison investigating whether Large Reasoning Models (LRMs) are superior judges to non-reasoning LLMs. O
arXiv:2603.12554v2 Announce Type: replace-cross Abstract: Reinforcement learning (RL) has been effective for post-training autoregressive (AR) language models, but extending these methods to diffusion
arXiv:2605.14366v1 Announce Type: new Abstract: Extending large language models (LLMs) to low-resource languages often incurs an 'alignment tax': improvements in the target language come at the cost o
arXiv:2605.14558v1 Announce Type: cross Abstract: Agentic reinforcement learning trains large language models using multi-turn trajectories that interleave long reasoning traces with short environment
arXiv:2512.21651v2 Announce Type: replace Abstract: Large Language Models (LLMs) deliver strong performance across a wide range of NLP tasks, but their massive sizes hinder deployment on resource-cons
arXiv:2511.13026v3 Announce Type: replace Abstract: Self-reflection mechanisms that rely on purely text-based rethinking processes perform well in most multimodal tasks. However, when directly applied
arXiv:2605.14497v1 Announce Type: cross Abstract: Offline-to-online reinforcement learning harnesses the stability of offline pretraining and the flexibility of online fine-tuning. A key challenge lie
arXiv:2605.14925v1 Announce Type: new Abstract: Drone-view geo-localization aims to match a query drone image, often captured under adverse weather conditions (e.g., rain, snow, fog), against a galler
arXiv:2605.14174v1 Announce Type: new Abstract: Safe navigation for mobile robots demands policies that remain reliable under the high-consequence perception uncertainty of cluttered environments. Yet
arXiv:2605.14982v1 Announce Type: cross Abstract: We address the discounted reward setting in reinforcement learning (RL). To mitigate the value approximation challenges in policy gradient methods, ac
arXiv:2605.14746v1 Announce Type: new Abstract: While large language models (LLMs) are trained to align with human values, their generations may still violate safety constraints. A growing line of wor