Step-3.7-Flash news

VoRTeC: Taming Foundation Flow for One-step Real time Video Compression

arXiv:2609.02291v2 Announce Type: replace-cross Abstract: Ultra-low bitrate video compression still faces critical challenges: traditional neural video compression inevitably introduces blurring artifacts, while diffusion-based generative video compression suffers from excessive decoding latency and

Counterfactual Fairness Audits of Multi-Step Clinical LLM Agents Require a Measured Per-Action Instability Floor

arXiv:2609.03221v1 Announce Type: new Abstract: Counterfactual audits are the standard tool for checking whether a clinical agent treats demographically distinct but clinically identical patients differently. They report a flip rate: how often an action changes when only the patient descriptor chang

Decoupled Analysis-Judging: An Automated Creativity Evaluator Using LLMs in Complex Multi-step Creativity Tasks

arXiv:2609.03432v1 Announce Type: new Abstract: Automated evaluation of creativity tasks remains challenging for LLM-as-a-Judge, as LLM is susceptible to biases such as verbosity bias and leniency bias. Such limitations are particularly evident in Contextually-Grounded and Procedurally-Structured Ta

Multi-step Proximal Policy Improvement in Offline Reinforcement Learning

arXiv:2609.03842v1 Announce Type: new Abstract: Offline reinforcement learning (RL) must reconcile two competing requirements: policy updates should stay near dataset-supported actions to keep value estimates reliable, yet meaningful gains often require moving beyond the behavior distribution. We de

Time Without Timesteps: Simulating Coupled Dynamical Systems via Self-Consistency

arXiv:2609.03358v1 Announce Type: new Abstract: Numerical simulation of dynamical systems is usually organized as a causal march through time: each state is computed from the previous one. We explore a different formulation for coupled systems. For each subsystem type we train a neural surrogate map

Projected Riemannian Gradient Descent for the Bures-Wasserstein Barycenter: Dimension-Independent Linear Convergence at Unit Step Size

arXiv:2609.03762v1 Announce Type: new Abstract: The computation of the Bures-Wasserstein (BW) barycenter of an ensemble of positive definite matrices arises throughout machine learning, optimal transport, and quantum information. Riemannian gradient descent (RGD) at unit step size -- the fixed-point

From Relaxed Indexability to Exact Indexability: A $t$-Step Approach for Partially Observable Restless Bandits

arXiv:2608.24167v2 Announce Type: replace Abstract: Whittle index policies offer a scalable method for restless multi-armed bandits, but under partial observability even determining the indifference subsidy at a single belief requires solving an infinite-horizon belief-state problem with no closed-f

Margins, Not Windows: Training-Free Per-Step Lossy Speculative Decoding

arXiv:2609.02897v1 Announce Type: new Abstract: Speculative decoding accelerates LLM inference by drafting candidate tokens and verifying them in parallel. Tree-attention drafters such as EAGLE-3 are widely adopted, yet typically hold two decisions fixed: (1) a strict token-match verification rule a

ROMS-IMLE: A Minimalist Approach to Competitive Single-Step Generative Modelling

arXiv:2607.19332v2 Announce Type: replace Abstract: Generative models have undergone many generations of evolution, from VAEs/GANs to diffusion/flow matching. Along the way, the underlying techniques have become more complicated and various beliefs about what drives strong empirical performance have

Simplify to Amplify: Achieving Information-Theoretic Bounds with Fewer Steps in Spectral Community Detection

arXiv:2602.17104v3 Announce Type: replace-cross Abstract: We propose a streamlined spectral algorithm for community detection in the two-community stochastic block model (SBM) under constant edge density assumptions. By reducing algorithmic complexity through the elimination of non-essential preproc

PRO-Step: Step-level Process Reward Optimization for Retrieval-Augmented Generation

arXiv:2609.01658v1 Announce Type: cross Abstract: Retrieval-Augmented Generation enhances Large Language Models by grounding responses in external knowledge, but multi-hop reasoning remains vulnerable to error propagation, where early retrieval failures confound subsequent steps. Standard outcome-ba

LoRA-GA$^2$: Low Rank Adaptation with Multi-step Gradient Adaptive Alignment

arXiv:2608.19800v2 Announce Type: replace-cross Abstract: Low-Rank Adaptation (LoRA) is a prominent fine-tuning method for large models, achieving competitive performance with reduced memory overhead. However, a persistent performance gap remains between LoRA and full fine-tuning. Recent studies hav

CAT-Flow: Curvature-Adaptive sTeps for Flow Matching

arXiv:2609.01746v1 Announce Type: new Abstract: Flow Matching has emerged as a leading framework for generative modeling, powering state-of-the-art systems such as FLUX and Stable Diffusion 3.5. However, the iterative nature of its ODE-based sampling process creates a fundamental efficiency bottlene

GeoSPRINT: Geometric Redundancy-Aware Step Pruning for Inference in Diffusion Trajectories

arXiv:2609.02160v1 Announce Type: cross Abstract: Diffusion models achieve high sample quality but remain expensive at inference time because sampling requires many sequential neural function evaluations (NFEs). Existing acceleration methods either use fixed step-skipping schedules, adapt step sizes

Stepwise Think-Critique: Interleaved Reasoning and Self-Critique in a Single LLM

arXiv:2512.15662v4 Announce Type: replace Abstract: Human beings solve complex problems through critical thinking, where reasoning and evaluation are intertwined to converge toward correct solutions. However, most existing large language models (LLMs) treat the reasoning and verification as separate

DLM-One: Diffusion Language Models for One-Step Sequence Generation

arXiv:2506.00290v2 Announce Type: replace Abstract: This paper introduces DLM-One, a score-distillation-based framework for one-step sequence generation with continuous diffusion language models (DLMs). DLM-One eliminates iterative refinement by aligning the scores of a student model's outputs with

Gradient Descent on Logistic Regression with Non-Separable Data and Large Step Sizes

arXiv:2406.05033v3 Announce Type: replace Abstract: We study gradient descent (GD) dynamics on logistic regression problems with large, constant step sizes. For linearly-separable data, it is known that GD converges to the minimizer with arbitrarily large step sizes, a property which no longer holds

Monitoring Web Agents Without Internal Signals: Observable Trajectories and Key-Step Supervision

arXiv:2609.02057v1 Announce Type: new Abstract: Reliable web-agent monitoring is difficult when model-internal uncertainty signals such as token logits are unavailable. In this work, we study prefix-level risk prediction for web agents using observable trajectory signals: given an evolving prefix, e

Multi-Mask Diffusion Language Models for Few-Step Generation

arXiv:2607.19686v3 Announce Type: replace Abstract: Masked diffusion models (MDMs) are a promising family of language generators, but achieving high-quality few-step generation remains challenging. In MDMs, all forward trajectories collapse to a single fully masked state, leaving no terminal entropy

Edit Knowledge, Not Just Facts via Multi-Step Reasoning over Background Stories

arXiv:2602.02028v3 Announce Type: replace Abstract: Enabling artificial intelligence systems, particularly large language models, to update knowledge and flexibly apply it during reasoning remains a central challenge. Existing knowledge editing approaches emphasize atomic facts, improving factual re

Fine-tuning a 350M Model for Better Structured Outputs in 100 GRPO Steps

Multi-Step Knowledge Interaction Analysis via Rank-2 Subspace Disentanglement

arXiv:2511.01706v3 Announce Type: replace-cross Abstract: Natural Language Explanations (NLEs) describe how Large Language Models (LLMs) make decisions by drawing on external Context Knowledge (CK) and Parametric Knowledge (PK). Understanding the interaction between these sources is key to assessing

ReDeck: Step-Level Render-Grounded Refinement for Document-to-Slide Generation

arXiv:2609.00194v1 Announce Type: new Abstract: Document-to-slide generation is challenging because slides are dense editable artifacts that require both faithful content selection and precise spatial layout. Recent slide agents adopt iterative reflection, but typically follow a monolithic "one vers

Beyond Token Positions: Safety Alignment Across Denoising Steps in Diffusion Language Models

arXiv:2609.00495v1 Announce Type: cross Abstract: Diffusion large language models (dLLMs) generate text through iterative denoising rather than left-to-right decoding. This generation paradigm introduces two axes that can influence safety alignment: when tokens are generated during denoising and whe

Diffusion as a Training Curriculum for Timestep-Free Iterative Reasoning

arXiv:2609.01449v1 Announce Type: new Abstract: Diffusion models and recursive reasoners are both iterative, but they carry information across iterations differently. We add a persistent hidden state to a diffusion denoiser and remove its timestep conditioning, leaving a single shared update that ca

Reward-guided Fine-Tuning of One-Step Generative Models via Wasserstein Gradient Flow

arXiv:2608.29647v1 Announce Type: new Abstract: To mitigate the time complexity of generative models, one-step generative models have recently emerged through direct mapping from noise to data in a single forward pass. However, the reward-guided fine-tuning method of one-step generative models remai

Reinforcing Step-level Reasoning for Effective Self-Correction in LLMs

arXiv:2608.11573v2 Announce Type: replace-cross Abstract: Achieving effective self-correction, where models verify and correct their own mistakes, remains a fundamental challenge for large language models (LLMs). In this work, we propose Self-Fix Step-DPO (SFS-DPO), a reinforcement learning based, t

A rigor-matched audit of periodic-step layer skipping for efficient llm inference: conflayers versus swift, with a supplemental analysis of trained routing alternatives

arXiv:2608.28846v1 Announce Type: cross Abstract: Layer-skipping methods for efficient LLM inference decide, at some granularity, which transformer layers to execute for a given input. We present a rigor-matched, three-seed audit of two periodic-step, search-based methods that make this decision onl

A Few Teacher Steps Go a Long Way: Cost-Efficient On-Policy Data Augmentation for Agent Post-Training

arXiv:2607.04574v2 Announce Type: replace-cross Abstract: For LLM agents, supervised fine-tuning is not only about teacher labels' quality, but also about which interaction contexts those labels condition on. Pure behavioral cloning uses full teacher demonstrations, creating a mismatch between teach

Last Step Matters: Early Uncertainty Cannot Predict Failure in Long-Horizon Agents

arXiv:2608.29685v1 Announce Type: new Abstract: Early failure prediction is important for long-horizon agents, as it enables timely intervention and can reduce inference and tool-use costs. Uncertainty quantification, such as verbal confidence and perplexity, offers a promising approach to detecting

Watch your steps: Dormant Adversarial Behaviors that Activate upon LLM Finetuning

arXiv:2505.16567v4 Announce Type: replace-cross Abstract: Finetuning open-weight Large Language Models (LLMs) is standard practice for achieving task-specific performance improvements. Until now, finetuning has been regarded as a controlled and secure process in which training on benign datasets lea

Multi-Step Forecasting of Grape Berry Temperature based on LSTM Model with Feed-Forward Attention

arXiv:2608.29008v1 Announce Type: new Abstract: Accurate forecasting of grape berry temperature (Tb) is essential for enabling timely heat stress management in vineyards. In this study, a feed-forward attention mechanism integrated with a Long Short-Term Memory network (FAM-LSTM) was developed and e

Learning PDE Time-Stepping with Neural Cellular Automata

arXiv:2608.30328v1 Announce Type: new Abstract: Classical numerical solvers for partial differential equations (PDEs) are computationally expensive to solve repeatedly across varying initial conditions, motivating the need for learned surrogates. In this paper, we propose a trainable Neural Cellular

Three Steps at a Time: Learning Representations from Action Sequences in Contrastive RL

arXiv:2608.30640v1 Announce Type: new Abstract: While self-supervised approaches to reinforcement learning have achieved strong results by learning representations of states and actions, a key open question is the time scale over which actions should be modeled. Departing from the standard formulati

Beyond the Answer Key: Robustness Evaluation of Large Language Models for Step-Level Mathematical Verification

arXiv:2608.28725v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly used as graders, verifiers, and process auditors, but most mathematical evaluations still emphasize final-answer accuracy. This can obscure whether a model can verify a non-canonical but valid solution trac

LiteSearch-VL: Small Multimodal Search Agents via Trajectory Distillation and Synthetic Step-DPO

arXiv:2608.29357v1 Announce Type: new Abstract: Multimodal search agents answer visual questions by interleaving image understanding, web retrieval, tool use, and evidence synthesis. Strong systems exist, but in two expensive regimes: proprietary frontier models such as GPT-5 and Gemini, or large op

Reviving our data foundations is the most disruptive step to data maturity

arXiv:2608.29368v1 Announce Type: new Abstract: The most disruptive step that enterprises of small-medium size and maturity can take to make the most of the latest technological advances in AI is to step back from the hype and focus on establishing or reviving a good knowledge foundation layer. It i

AdaVLA: Adaptive Step Flow Matching for Training-free Acceleration of Vision-Language-Action Models

arXiv:2608.29208v1 Announce Type: cross Abstract: Vision-Language-Action (VLA) models, built upon Vision-Language Models (VLMs), have significantly enhanced robotic capabilities by leveraging internet-scale knowledge and multimodal reasoning. However, the intensive computational overhead of VLAs con

AgenticRag-R1: Agentic Reinforcement Learning with Stack Memory for Multi-Step Reasoning, Retrieval and Memorizing

arXiv:2608.29622v1 Announce Type: cross Abstract: Retrieval-Augmented Generation (RAG) improves the factuality of large language models (LLMs), yet existing RAG systems often struggle with complex, multi-step reasoning that requires adaptive retrieval and continuous revision of intermediate contexts

SegBench-GC: Testing Segmentation Invariance in Multi-Step Offline Goal-Conditioned Reinforcement Learning

arXiv:2608.27678v1 Announce Type: new Abstract: Offline goal-conditioned reinforcement learning (GCRL) often uses trajectory structure for future-goal sampling and multi-step targets, yet logged trajectories may be partitioned for administrative reasons that do not correspond to termination. We intr

SafeStep: An Interactive Demonstration of Semantic Communication for Pedestrian Safety Monitoring

arXiv:2608.27688v1 Announce Type: new Abstract: In this paper, we develop SafeStep, an interactive browser-based semantic communication platform for live pedestrian safety monitoring. SafeStep extracts pedestrian information from four live traffic-camera feeds, transmits it through a semantic commun

STEP: State-Aware Task Estimation and Planning with Multi-Modal LLMs for Human-Robot Collaboration

arXiv:2608.27225v1 Announce Type: cross Abstract: Effective human-robot collaboration in industrial settings requires robots to understand human intentions and assist with task planning, reducing workload. Recent works have explored the use of Multi-modal Large Language Models (MM-LLMs) for task pla

TRACE-CRC: Trajectory-Adaptive Conformal Risk Control for Multi-Step Channel State Information Prediction

arXiv:2608.27124v1 Announce Type: new Abstract: Reliable prediction of time-varying channel state information (CSI) is essential for efficient wireless communication. Each CSI frame is a matrix-valued representation of the wireless channel response, and a sequence of CSI frames forms a temporal chan

TRACES: Tagging Reasoning Steps for Adaptive Cost-Efficient Early-Stopping

arXiv:2604.21057v2 Announce Type: replace Abstract: The field of Language Reasoning Models (LRMs) has been very active over the past few years with advances in training and inference techniques enabling LRMs to reason longer, and more accurately. However, a growing body of studies show that LRMs are

StepGuard: Learning Step-Level Guardrails with Scalable Supervision and Safety-Utility Balancing

arXiv:2608.24777v1 Announce Type: new Abstract: LLM-based agents can interact with external environments through tool invocation, but this capability also introduces security risks such as file modification, information leakage, and unauthorized actions. Existing guardrails often evaluate completed

Omanic: Towards Step-wise Evaluation of Multi-hop Reasoning in Large Language Models

arXiv:2603.16654v3 Announce Type: replace-cross Abstract: Evaluating the reasoning abilities of large language models (LLMs) solely from final answers can obscure failures in intermediate steps, especially in multi-hop QA benchmarks without step-level annotations. To address this gap, we introduce O

Credit Without Ground Truth: Auditing Step-Level Credit Assignment in LLM Agents Against Executed Replay

arXiv:2608.19760v2 Announce Type: replace-cross Abstract: Audited against policy-conditional ground truth from executed replay in a single-agent tool environment (ALFWorld), none of the step-level credit signals we audit -- LLM-judge scores, outcome-conditioned logprob ratios, or the policy's own co

Modeling spatio-temporal locality in multi-step forecasting of geo-referenced time series

arXiv:2608.25698v1 Announce Type: new Abstract: Forecasting future measurements from geographically distributed sensors is essential across many domains. However, the spatial distribution of these sensors raises multiple challenges, primarily due to spatial autocorrelation phenomena, that introduce

Automata from Agent Traces: Failure and Next-Step Prediction

arXiv:2608.23670v1 Announce Type: new Abstract: LLM-based agents execute multi-step tasks, but their behavioral structure remains opaque: long unstructured traces resist the safety auditing and runtime monitoring that deployment requires. Existing approaches operate per-trace or success-only, so the