Artificial Intelligence
Artificial Intelligence news — Thursday 15 October 2026
150 stories from 11 sources, every one in English. Each links to its publisher.
-
Granite Switch 4.1: Chat, Decisions, and Speech from One Endpoint
Hugging Face Blog
-
Bringing GPT-Rosalind to academic research labs
OpenAI Blog
Introducing the Rosalind Academic Plan, offering eligible academic labs sponsored access to GPT-Rosalind and tools to advance life sciences research.
-
SpaceXSI, the highest performance ‘Glock Voice 3.0’ released... “Secure Phone Resolution No. 1”
AI Times (Korea)
Translated from Korean
-
DeepCock Low-cost technology, performance in accordance with the information location, simplicity exposure
AI Times (Korea)
Translated from Korean
-
Liquid AI, Flagship Model 'LFM2.5-24B-A2B' Released..."OnDevice Agent aimed at"
AI Times (Korea)
Translated from Korean
-
‘Melon’ A2O Lite, AI Avatar
AI Times (Korea)
Translated from Korean
-
Tencent Agent also bypasses security network, Alababa map service Mass scan and data extraction
AI Times (Korea)
Translated from Korean
-
[Bulletin] OpenAI Korea, ‘Devday Explorer: Seoul’
AI Times (Korea)
Translated from Korean
-
US 8 states, openAI relative group litigation... “ChatGPT self-sufficient”
AI Times (Korea)
Translated from Korean
-
OpenAI·Antropicture "Different Chinese Open Model, Hole on Cyber Security" Warning
AI Times (Korea)
Translated from Korean
-
MS, 2~6 weeks weather forecast Blank Mew opens new AI framework
AI Times (Korea)
Translated from Korean
-
"AI agent hack, early warning to strengthen system security"
AI Times (Korea)
Translated from Korean
-
Accelerating the search for antimicrobial peptides with Google Antigravity
Google DeepMind
Accelerating the search for novel antimicrobial peptides using Google Antigravity for Science.
-
Launched ‘Beads’, a AI platform, aimed at individuals beyond the university and public
AI Times (Korea)
Translated from Korean
-
1182 elements inspection, street light, security, etc.
AI Times (Korea)
Translated from Korean
-
Jeonnam-Gwangju Semiconductor cluster, government and corporate site inspection for the first mass production goal in 2030
AI Times (Korea)
Translated from Korean
-
Vishai, 50 times a year sales.. .2Based GPU Loan Promotion
AI Times (Korea)
Translated from Korean
-
Real search on Bear Photo made by ChatGPT...Russian fake image diffuser fine
AI Times (Korea)
Translated from Korean
-
Jev Your Engines: Seven Open Decision Models vs. a Generative LLM
Towards Data Science
A hands-on Jev repo comparison of routing accuracy, latency, and output reliability The post Jev Your Engines: Seven Open Decision Models vs. a Generative LLM appeared first on Towards Data Science .
-
Over $10,000 for the first time learning, ‘Payable 5’... ‘Ultramarine 1 Flash’ released
AI Times (Korea)
Translated from Korean
-
Open From Data Center to Desktop: NVIDIA Expands Nemotron Family With Highest Efficiency Super VL and Bolt Models
NVIDIA AI Blog
NVIDIA today released Nemotron 3.5 Super VL, an open 120-billion-parameter model that delivers the fastest task completion and highest accuracy for specialized always-on agents. Joining it are two new edge models: Nemotron 3.5 Bolt 74B-A9B…
-
OpenAI, the infrastructure leader, replaces...Datacenter direct securing speed
AI Times (Korea)
Translated from Korean
-
Antropic, started IPO roadshow...exported 45%, but expecting an investor
AI Times (Korea)
Translated from Korean
-
What OpenAI’s 719 Math Papers Mean for Data Science
Towards Data Science
Matching messy text, choosing representative examples, and generating comparison data. The post What OpenAI’s 719 Math Papers Mean for Data Science appeared first on Towards Data Science .
-
U.S. large-scale photovoltaic expansion, associated with the reduction of electricity bills...
AI Times (Korea)
Translated from Korean
-
Indulge with a large-scale submarine contract, 120 pieces of corporate value
AI Times (Korea)
Translated from Korean
-
Automated solar construction logistics...Release, AFR for autonomous driving
AI Times (Korea)
Translated from Korean
-
Apple, December Chinese AI first line...Ariba·Baidu Handhold in half a year
AI Times (Korea)
Translated from Korean
-
Teaching students to think and learn through questions
OpenAI Blog
Celebrate Teachers’ Day with AI for Educators in Brazilian Portuguese and see how Study Mode can help teachers support active, step-by-step learning.
-
Italy Solar, 2030 82.3GW...2040 Max 136.7GW View
AI Times (Korea)
Translated from Korean
-
Jenson Huang “Unlimited AI developments can suspend chip supply”
AI Times (Korea)
Translated from Korean
-
How Many Guesses Does It Take to Crack an AI's "Random" Number?
Towards Data Science
I read the exact answer distribution from the logits of open models, base and instruction-tuned, and turned it into bits and guesses-to-crack. The post How Many Guesses Does It Take to Crack an AI's "Random" Number? appeared first on…
-
The Download: reviving frozen pig kidneys and boosting nuclear power
MIT Technology Review Biotech
This is today’s edition of The Download , our weekday newsletter that provides a daily dose of what’s going on in the world of technology. Pig kidneys that were “frozen” for a year have been brought back to life Pig kidneys that were…
-
OpenAI Docheng Anaerobic Engineer, Saudi.. "There was also concerns"
AI Times (Korea)
Translated from Korean
-
OpenAI, Apple’s... "All employees are indispensable to unopened devices"
AI Times (Korea)
Translated from Korean
-
VNTG·Korea MS, co-built manufacturing site AI transition execution system
AI Times (Korea)
Translated from Korean
-
I Tested Four Ways to Give an AI Assistant a Memory. The Obvious One Was the Worst Value.
Towards Data Science
Pasting in the whole chat history cost 20 to 31 times more than looking up notes, and it did no better on the tasks. The post I Tested Four Ways to Give an AI Assistant a Memory. The Obvious One Was the Worst Value. appeared first on…
-
How much more can nuclear power plants give?
MIT Technology Review Biotech
In this age of rising electricity demand, people are trying to squeeze every last bit of power out of every power plant possible. Over the past few weeks, there’s been big news about nuclear power plants trying to do just that…
-
BADRAFT, NPU-only language model pre-learning success...GPU-dependent low
AI Times (Korea)
Translated from Korean
-
Apertus 1.5 Release Notes
Hugging Face Blog
-
Apertus 1.5 Release Notes
Hugging Face Blog
-
Asus expands the domestic market of ‘Agent Computer’... From AI PC to AI for business
AI Times (Korea)
Translated from Korean
-
Six wild ideas for perpetuating your mind
MIT Technology Review Biotech
Have you ever thought about what makes you you ? Science assures us it has everything to do with the three-pound organ between our ears, with its billions of neurons and trillions of delicate connections. But it’s a lot trickier to explain…
-
[AIReader’s Book] Vibe Coding Production Principle
AI Times (Korea)
Translated from Korean
-
Pig kidneys that were “frozen” for a year have been brought back to life
MIT Technology Review Biotech
Pig kidneys that were cooled to -150 °C and stored in a freezer for months have been successfully rewarmed and shown to work again for the first time, say scientists. They hope to replicate the findings with human organs. The kidneys—which…
-
[Oct.14] The reason for the character dotAI that has been immersed with chatGPT in four years
AI Times (Korea)
Translated from Korean
-
Computational Reflective Equilibrium for Agentic AI: A Rawlsian Alignment Scaffold for Governance
arXiv cs.AI
arXiv:2610.16015v1 Announce Type: new Abstract: The rapid deployment of agentic AI has introduced complex challenges for the ethics and governance of AI systems. Agents built on large language models act through tools on behalf of people…
-
Same Illness, Different Voice: Restating the Patient's Account Narrows Narrative Anchoring in Clinical Language Models
arXiv cs.AI
arXiv:2607.27384v2 Announce Type: replace-cross Abstract: Clinical language models are evaluated on vignettes written in a clinician's voice, but patients describe the same illness in their own. We pair 1,000 MedQA cases with three…
-
Beyond Size: How Gradients Shape Pruning Decisions in Large Language Models
arXiv cs.CL
arXiv:2311.04902v3 Announce Type: replace Abstract: Large Language Models (LLMs) with billions of parameters are prime targets for network pruning, which removes some model weights without hurting performance. Prior approaches such as…
-
ScholarQuest: A Taxonomy-Guided Benchmark for Agentic Academic Paper Search in Open Literature Environments
arXiv cs.AI
arXiv:2606.20235v2 Announce Type: replace-cross Abstract: Academic paper search is a core step in scientific research, and LLM-based search agents are emerging as a promising paradigm for iterative, intent-driven literature exploration…
-
TreeFlex: Agents Navigating Their Own Search Trees
arXiv cs.CL
arXiv:2610.20334v1 Announce Type: cross Abstract: A ReAct-style agent accumulates observations, intermediate results, and failed attempts in a single context. As this context grows, stale and off-path information consumes its limited…
-
Similar Models, Broken Odds: Where Luce's Axiom Fails in Chatbot Arena
arXiv cs.LG
arXiv:2610.19095v1 Announce Type: cross Abstract: Chatbot Arena ranks language models with the Bradley-Terry model, which gives each model one score and so assumes Luce's choice axiom. Choice theory predicts where this fails: near-copies…
-
A Sovereign, Indic-First, Reasoning Model
arXiv cs.AI
arXiv:2610.15371v1 Announce Type: cross Abstract: Gnani-Evon-v3.3-30B-A3B is a long-context hybrid Mamba-Transformer mixture-of-experts model for reasoning, instruction following, and tool use in English and eight Indian languages…
-
MemoryVLN: From Ego-Aware Memory Representation to Cross-Context Memory Reuse in Vision-Language Navigation
arXiv cs.AI
arXiv:2610.18478v1 Announce Type: cross Abstract: Large vision-language models (LVLMs) have shown strong potential for embodied vision-language navigation (VLN), yet their performance degrades on long-horizon trajectories due to…
-
Latent Compaction with Test-Time Training
arXiv cs.LG
arXiv:2610.20582v1 Announce Type: cross Abstract: Long contexts enlarge key--value (KV) caches, increasing memory use and attention costs during generation. Compressing these caches must preserve information for unseen questions. We…
-
APEX: Assumption-free Projection-based Embedding eXamination Metric for Generative Distribution Evaluation
arXiv cs.AI
arXiv:2605.07786v4 Announce Type: replace-cross Abstract: As generative models achieve unprecedented visual quality, the gold standard for image evaluation remains traditional feature-distribution metrics. However, these metrics are…
-
Revealing the Gap in Human and VLM Scene Description through Counterfactual Semantic Saliency
arXiv cs.AI
arXiv:2605.13047v2 Announce Type: replace-cross Abstract: Evaluating whether large vision-language models (VLMs) align with human perception for high-level semantic scene comprehension remains a challenge. Traditional white-box…
-
Never Too Late for Force: Accelerating VLA Post-Training with Reactive Force Injection
arXiv cs.AI
arXiv:2607.14236v3 Announce Type: replace-cross Abstract: Pretrained vision-language-action (VLA) policies provide strong language-conditioned manipulation knowledge, but they remain largely vision-driven and can struggle once manipulation…
-
Revocation Is Not Complete at Receipt: An Executable Outcome Contract for Distributed Agent-Tool Gateways
arXiv cs.AI
arXiv:2610.15991v1 Announce Type: new Abstract: Distributed agent workflows cache delegated authority at tool gateways. Revocation may be reported at issuer commit, transport acknowledgment, or receiver persistence, although none of these…
-
The Dice Roll Method: A Standardized Protocol for Repeated-Query Auditing of Large Language Model Brand Recommendations
arXiv cs.CL
arXiv:2609.04047v2 Announce Type: replace-cross Abstract: Background: Researchers increasingly use repeated identical prompts to audit stochastic variation in large language model (LLM) brand recommendations, yet no standardized protocol…
-
CAST: History-Adaptive Spectral Clipping for LLM Pretraining
arXiv cs.LG
arXiv:2610.19761v1 Announce Type: new Abstract: Soft spectral clipping (SSC) reduces gradient singular values, but its effect depends on the gradient scale and clipping threshold. We introduce CAST (Clipping with Adaptive Scales and…
-
Self-Retrospection Distillation: Turning Post-hoc Experiences into Prior Foresight
arXiv cs.CL
arXiv:2610.08077v3 Announce Type: replace-cross Abstract: Reinforcement learning with verifiable rewards (RLVR) turns agent experience into learning signals primarily through scalar outcome rewards after interaction. For group-relative…
-
Scaling Vision Transformers Without Training: Prune-and-Merge Token Compression
arXiv cs.AI
arXiv:2610.18039v2 Announce Type: cross Abstract: Vision Transformers achieve strong performance but incur high computational costs from quadratic token interactions. Training-free token reduction methods improve efficiency, but their…
-
Agaram 2.5B: The First Tamil-First Language Model Trained from Scratch with Tool Calling. A Playbook for Under-Represented Languages
arXiv cs.AI
arXiv:2610.18420v1 Announce Type: cross Abstract: We trained Agaram 2.5B, a 2.5-billion-parameter Tamil-English language model, from scratch on about 50 billion tokens (58% Tamil, the rest English, math, code and parallel text) on one node…
-
Memory by Design: Bayesian Layers for Sequence Modeling
arXiv cs.LG
arXiv:2605.31163v5 Announce Type: replace-cross Abstract: We introduce the \emph{design-model framework}: a way to derive efficient recurrent sequence maps from explicit assumptions about memory. A design model writes evidence into memory…
-
PatTransBench: an ipc-stratified, multi-dimensional benchmark for patent machine translation evaluation
arXiv cs.AI
arXiv:2610.16538v1 Announce Type: cross Abstract: We present PatTransBench, a 12,000-item Chinese-English/Japanese/Korean patent machine translation (PatMT) benchmark with a four-dimensional evaluation protocol. The benchmark comprises…
-
Tev: Scaling Long-Horizon Agentic Environments with Dense Progress Rewards
arXiv cs.AI
arXiv:2610.18868v1 Announce Type: cross Abstract: Long-horizon agents have drawn growing attention for their ability to work autonomously -- from hours to days -- on economically valuable tasks, a capability essential to automated…
-
Reading the Room: Foundations, Design, and Challenges of Normative Competence in LLMs
arXiv cs.AI
arXiv:2610.10906v2 Announce Type: replace Abstract: Human communities are governed by normative systems: shared standards that produce \textit{norms} dictating acceptable behavior, enforced through community sanctioning. Aligning…
-
Budgeted KV-Cache Repair for Evolving Contexts
arXiv cs.AI
arXiv:2610.17338v1 Announce Type: new Abstract: Language-model agents revise their working context as information becomes obsolete, but deleting text does not erase its influence on cached computation. Later key-value (KV) states remain…
-
When and Why Adversarial Training Improves PINNs: A Neural Tangent Kernel Perspective
arXiv cs.LG
arXiv:2605.15959v2 Announce Type: replace Abstract: Physics-informed neural networks (PINNs) are powerful surrogates for differential equations but are notoriously difficult to train due to spectral bias, stiffness, and poor accuracy on…
-
LiFUSE: Frequency-Spatial Unified Self-Supervised Pretraining for Liver Disease Analysis on Non-contrast CT
arXiv cs.AI
arXiv:2610.18243v1 Announce Type: cross Abstract: Given the growing global burden of liver tumors and diffuse liver diseases, non-contrast computed tomography (NCCT) is increasingly used for opportunistic screening because of its low cost…
-
Posterior Frequencies and the Cost of Not Fitting: A Correction to a Memorization Lower Bound
arXiv cs.LG
arXiv:2610.20402v1 Announce Type: cross Abstract: Feldman's memorization lower bound assigns a frequency coefficient based on how often a point appears in the training sample. We give an exact three-point counterexample to its conditional…
-
Resource-Efficient 3D Shape and Orientation Estimation of Road Users using Roadside RGB Camera and 3D Radar
arXiv cs.AI
arXiv:2610.18417v1 Announce Type: cross Abstract: End-to-end deep learning pipelines for camera-radar 3D object detection are well established in driver assistance systems and autonomous vehicles; however, adapting them to roadside-mounted…
-
Learning to Select Features with Manifold Deep Unfolding PCA
arXiv cs.LG
arXiv:2610.19436v1 Announce Type: cross Abstract: Unsupervised feature selection (UFS) is of great significance in image processing and pattern recognition. Nevertheless, existing model-based methods often rely on manually tuned…
-
AutoSec-RAN: Specification-Grounded LLM Fine-Tuning for Automated Security Analysis of O-RAN
arXiv cs.AI
arXiv:2610.15995v1 Announce Type: cross Abstract: This paper presents AutoSec-RAN, a specification-grounded large language model (LLM) framework for automated security triage of Open Radio Access Network (O-RAN) testing. Unlike approaches…
-
When and How Order-Flow Interventions Work After Dealers Adapt
arXiv cs.AI
arXiv:2610.18158v1 Announce Type: new Abstract: Learning dealer agents can settle on persistently wide quotes, raising what customers pay to trade. A venue can route more orders to tighter quotes, but a wide quote may reflect real risk…
-
ProximalFM: Amortized Proximal Causal Inference under Hidden Confounding
arXiv cs.LG
arXiv:2610.08078v2 Announce Type: replace-cross Abstract: Standard causal identification methods often assume no unmeasured confounding and can fail when relevant confounders are unobserved. Proximal causal inference instead uses proxy…
-
Manac\'a-1B-Instruct-2: Reproducible Safety Alignment for a Native Brazilian-Portuguese LLM. Negative DPO Results, a Controlled Composition Scan, and a Silent Failure Mode of the LLM Judge
arXiv cs.CL
arXiv:2610.20570v1 Announce Type: new Abstract: Brazilian Portuguese remains underserved by open-source language models, and the few that are safety-aligned rarely document what did not work or validate comparisons with statistical…
-
Fully sequential Gaussian process regression for preference learning
arXiv cs.LG
arXiv:2610.19391v1 Announce Type: new Abstract: We study continuous global optimization problems in which the objective function cannot be directly evaluated; rather, we observe only which of two selected design points appears to perform…
-
CoverSynth: Coverage-Guided Generation of Retrieval Evaluation Sets
arXiv cs.AI
arXiv:2610.15409v1 Announce Type: new Abstract: Enterprise agents need retrieval evaluation sets that expose domain-specific failures and guide developers towards efficient deployment tradeoffs. Neither expert-human-authored sets nor…
-
DualOPSD: Closing the Feedback Loop in On-Policy Self-Distillation
arXiv cs.AI
arXiv:2608.26019v2 Announce Type: replace-cross Abstract: On-policy self-distillation (OPSD) combines student generated trajectories with dense supervision from a privileged teacher of the same model. Its pointwise clipping protects the…
-
One Block Ahead: Promoting Sampler State in Diffusion Language Models
arXiv cs.CL
arXiv:2610.19510v1 Announce Type: new Abstract: Block-diffusion language models denoise a whole block of token positions in parallel so each forward pass can emit many tokens. However, blocks are still denoised in strict sequence, a cost…
-
A Decade of PhotoMappers: A Longitudinal Study of Crowdsourced Disaster Photos Geolocalization with Vision-Language Models and Geospatial Reasoning
arXiv cs.AI
arXiv:2610.15403v1 Announce Type: new Abstract: Crowdsourced disaster imagery provides timely, ground-level observations when conventional disaster mapping is delayed or incomplete, yet its geographic value depends on where observations…
-
SWE-Spot: Building Small Repo-Experts with Repository-Centric Learning
arXiv cs.LG
arXiv:2601.21649v2 Announce Type: replace Abstract: The deployment of coding agents in privacy-sensitive and resource-constrained environments drives the demand for capable open-weight Small Language Models (SLMs). However, they suffer…
-
Counterfactual Rollouts for Offline Contrastive Goal-Conditioned Reinforcement Learning
arXiv cs.AI
arXiv:2610.14953v1 Announce Type: cross Abstract: Contrastive reinforcement learning (CRL) learns goal-conditioned values by associating state--action pairs with the states subsequently reached, assuming that no unobserved variable jointly…
-
HACHIMI-II: From Role-Play to Learner Behavior Modeling for Faithful Student Simulation
arXiv cs.CL
arXiv:2610.19654v1 Announce Type: new Abstract: Simulated students let researchers test tutoring systems and teaching strategies before real learners are available, but only if the simulation reproduces where a particular learner succeeds…
-
SearchMaster: Grounded and Regulated Self-Play for Search Agents
arXiv cs.AI
arXiv:2608.01822v2 Announce Type: replace Abstract: Training LLM-based search agents requires high-quality search data: tasks that demand genuine multi-hop retrieval and trajectories that use search tools effectively. Existing pipelines…
-
Integrating AIRL-Based Reward Estimation into a Control Method Using Fuzzy Inference and Policy Gradient
arXiv cs.AI
arXiv:2610.14723v1 Announce Type: cross Abstract: A method integrating policy gradient with fuzzy inference has been proposed. This method, hereafter referred to as the base model, has high flexibility with respect to the configuration of…
-
Comparing GenAI to Machine Learning and Manual Review for eDiscovery
arXiv cs.AI
arXiv:2610.16925v1 Announce Type: cross Abstract: Keeling et al. [Keeling2026Generative] report on a case study recently conducted by Redgrave LLP in which they assert that Relativity's Generative AI Review tool (aiR) achieved 88% recall…
-
Test-Time Forecasting: Recurrent Depth Enables In-Context Dynamics Prediction
arXiv cs.LG
arXiv:2610.19552v1 Announce Type: new Abstract: Foundation models for dynamical systems can forecast unseen partial differential equations (PDEs), ordinary differential equations (ODEs), and time series during inference. However, it…
-
InstructPro2: Natural Language and Biological Grounding Guided De Novo Protein Design
arXiv cs.LG
arXiv:2610.19592v1 Announce Type: new Abstract: The ability to generate proteins with desired functions is a central goal of computational biology. Although natural language can express diverse design objectives, existing protein…
-
Harness^2: Recursive Agent Harnessing for an Open World
arXiv cs.AI
arXiv:2610.14224v1 Announce Type: new Abstract: LLM agents' capabilities depend not only on their underlying models but also on their execution harnesses. Recursive self-improvement (RSI) offers a promising approach to automatically evolve…
-
Get In, Decide, Prove: A Merchant-Side Readiness Benchmark for Agentic Commerce
arXiv cs.AI
arXiv:2610.17378v1 Announce Type: new Abstract: AI agents now browse and buy on people's behalf, and benchmarks measure how well the agents perform. Far less is known about the other side of the transaction: whether a merchant's website…
-
Style Over Substance: Content-Invariant Wrappers Flip LLM Safety-Judge Verdicts
arXiv cs.AI
arXiv:2609.08236v2 Announce Type: replace Abstract: Automatic safety judges -- for example Llama Guard, or GPT-4o given a grading prompt -- produce the numbers behind reported jailbreak success rates, defense evaluations, and safety…
-
Agronomic Characterization and Time-Series UAV Phenotyping to Support Field Pea Breeding in Iowa
arXiv cs.CV
arXiv:2610.19154v1 Announce Type: new Abstract: Field pea has gained interest in the Midwest U.S. and has the potential to benefit Iowa cropping systems. This study evaluated 17 accessions and 7 Canadian varieties for agronomic and seed…
-
Jev-ToM: Evaluating Theory of Mind in Typed Decision Models
arXiv cs.AI
arXiv:2610.18610v1 Announce Type: cross Abstract: Dialogue and collaborative assistants repeatedly judge what a partner knows, believes, or intends. Typed decision models offer a potential alternative to generative large language models…
-
GraphDecisionBench: Benchmarking System One Models on Graph Tasks
arXiv cs.AI
arXiv:2610.06354v3 Announce Type: replace Abstract: Large language models (LLMs) are increasingly explored for graph understanding and decision-making, while System One models such as Jev select directly from supplied options. However, the…
-
GameSmith: Scaling Zero-Sum Self-Play with Synthesised Games
arXiv cs.AI
arXiv:2610.13995v1 Announce Type: cross Abstract: Post-training large language models (LLMs) on two-player zero-sum games provides a verifiable reward, and the gains have been shown to transfer to general reasoning. Existing pipelines…
-
Batch Bayesian Active Learning with Partial Batch Label Sampling
arXiv cs.AI
arXiv:2510.09877v4 Announce Type: replace-cross Abstract: Over the past couple of decades, many active learning acquisition functions have been proposed, leaving practitioners with an unclear choice of which to use. Bayesian active…
-
EduBehaviors: Assertion-based Schemas for Auditable Coding of Educational Dialogues
arXiv cs.CL
arXiv:2609.27043v2 Announce Type: replace Abstract: Large language models have enabled rapid deployment of language classification and analysis, allowing users to describe complex constructs in natural language, such as pedagogical moves…
-
Beyond Aggregate Accuracy: Item-Level Transitions in Prompt Optimization
arXiv cs.LG
arXiv:2610.19084v1 Announce Type: new Abstract: Automatic prompt optimization (APO) uses feedback from model responses to improve prompts automatically. However, its effectiveness is usually measured through overall scores, which do not…
-
Activation Steering with Jungian Cognitive Functions
arXiv cs.AI
arXiv:2607.20803v3 Announce Type: replace-cross Abstract: Controlling combinations of personality-related tendencies is important for adapting large language models to diverse interaction needs. But successful single-trait steering does…
-
Divide, Consult, Conquer: Capability Laundering Through Aligned LLMs
arXiv cs.AI
arXiv:2609.15383v2 Announce Type: replace-cross Abstract: Language model safety is typically evaluated one interaction at a time. We show that a weaker, unaligned model can split a harmful task into benign-looking subproblems, consult a…
-
OctoLong: Mid-Training On Cross-Repository Code Contexts Enhances Long-Context Modeling
arXiv cs.AI
arXiv:2608.05141v2 Announce Type: replace Abstract: Context lengths of language models (LMs) have dramatically increased, driven by the demands for in-context learning, self-improvement, and long-horizon agentic workflows. However…
-
EmoStance: Response-Side Affective-Orientation Control for Empathetic Response Generation via Emoji Weak Supervision
arXiv cs.AI
arXiv:2609.02133v2 Announce Type: replace Abstract: Empathetic response generation requires models to decide not only what to say, but also how to respond to the previous speaker's affective situation. We formulate this as response-side…
-
When Do Learned Blocks Preserve Causal Meaning? Block Identifiability and Constructive Causal Abstractions
arXiv cs.AI
arXiv:2610.18155v1 Announce Type: cross Abstract: Standard identifiability theory assumes one-dimensional latent factors, but learned representations can encode latent information across multidimensional blocks whose internal coordinates…
-
TIMERSI: Towards Self-Improved Agentic Deployment for Time-Series Foundation Models
arXiv cs.LG
arXiv:2610.19725v1 Announce Type: new Abstract: Time-series foundation models provide strong zero-shot forecast, but deployment for real- world applications still requires task-specific designs for model inputs and outputs. Direct…
-
RAIL: Rethinking Auditory Intelligence in Large Audio-Language Models with a CHC-Grounded Benchmark
arXiv cs.AI
arXiv:2606.11260v2 Announce Type: replace-cross Abstract: Humans process rich auditory environments through various cognitive capabilities such as audio perception, audio reasoning and memory. Despite recent progress in large…
-
Real2SAM2Real: Generative 3D Caches as Complementary Context for Video Diffusion
arXiv cs.AI
arXiv:2606.00299v2 Announce Type: replace-cross Abstract: Video Diffusion Models (VDMs) synthesize high-quality videos, but precise camera and motion control remains challenging when large viewpoint shifts, changing occlusions, or complex…
-
Evaluating Large Language Models for Multilingual Abstractive Summarization: A Comprehensive Study of English and Bengali
arXiv cs.CL
arXiv:2610.19003v1 Announce Type: new Abstract: Abstractive summarization remains challenging across languages, where linguistic quality and factual consistency may diverge, particularly in low-resource settings. We present a standardized…
-
Wong--Zakai Convergence and FNO Learning for Stochastic Volterra Equations
arXiv cs.LG
arXiv:2610.18939v1 Announce Type: cross Abstract: This paper studies Wong--Zakai convergence and Fourier neural operator learning for stochastic Volterra integral equations (SVIEs) describing random systems with memory. A Wong--Zakai…
-
Fantastic Metrics and How to Find Them: Meta-Evaluating Metrics for Multi-lingual Style-Personalized Generation
arXiv cs.CL
arXiv:2610.19166v1 Announce Type: new Abstract: Meta-evaluation of text generation metrics and LLM-judges is primarily conducted using metric-human correlations. However, prior work has shown that correlation is not an appropriate estimate…
-
From Suboptimal to Optimal: One-Step Latent Flow Policies for Offline Reinforcement Learning
arXiv cs.AI
arXiv:2610.16984v1 Announce Type: cross Abstract: Offline reinforcement learning aims to learn effective policies from fixed datasets without environment interaction. Although generative models such as diffusion and flow matching enhance…
-
SPECTRA-VLA: Base-Fine Hyperspectral Token Transfer at Frozen Vision-Language-Action Interfaces
arXiv cs.CV
arXiv:2610.19651v1 Announce Type: new Abstract: Hyperspectral imaging distinguishes RGB-similar objects, but a pretrained vision-language-action model needs a state its frozen decoder can use. SPECTRA-VLA provides a native-width spectral…
-
A Diagnostic Study of World-Space Human Recovery Under High Camera Dynamics
arXiv cs.CV
arXiv:2610.19729v1 Announce Type: new Abstract: Monocular world-space human recovery methods excel on current benchmarks, but these rarely include fast camera motion. We introduce the K-Pop stress test, broadcast dance videos that combine…
-
Why 3D Reconstruction Breaks Down on Under-Ice AUV Video, and How to Predict It
arXiv cs.CV
arXiv:2610.19993v1 Announce Type: new Abstract: Reconstruction of 3D structures using under-ice AUV video footage has been inaccurate, although it is not certain whether this is due to shortcomings in the methods or an inherent deficiency…
-
When Pairwise Preferences Cycle: A Reliability-Anchored Preference Walk for Video GRPO
arXiv cs.CV
arXiv:2610.20089v1 Announce Type: new Abstract: Pairwise preference feedback has become a widely adopted supervision signal for reinforcement learning in video generation models, as pairwise comparisons provide a more reliable alternative…
-
DreamHMR: From Video Generation to Human Motion Modeling via Structural Focusing
arXiv cs.CV
arXiv:2610.20124v1 Announce Type: new Abstract: Human mesh recovery (HMR) provides a geometric foundation for understanding human motion, but accurately reconstructing both body motion and fine hand articulation from videos remains…
-
Vision-Language Models Integrate Conflicting Evidence Like Biased Bayesian Observers
arXiv cs.CV
arXiv:2610.19407v1 Announce Type: cross Abstract: Vision-language models must reconcile conflicting text and images. Prior work has characterized the resulting behavior in terms of modality biases--e.g., a "blind faith in text"--but why…
-
GenState-AI: State-Aware Dataset for Text-to-Video Retrieval on AI-Generated Videos
arXiv cs.CV
arXiv:2603.14426v2 Announce Type: replace Abstract: Existing text-to-video retrieval benchmarks often permit matching from static appearance, making it difficult to determine whether models capture state changes and final outcomes. We…
-
Efficient Noise Context-Aware Representation for Low-Dose CT Denoising
arXiv cs.CV
arXiv:2608.10343v2 Announce Type: replace Abstract: While deep learning-based denoising has become widely adopted in low-dose CT, conventional models use generic architectures designed for natural images, failing to account for…
-
RealityBench: Simulated Environments to Evaluate a Vision Language Model's Ability to Document Human Physical Work
arXiv cs.CV
arXiv:2610.14397v2 Announce Type: replace Abstract: Augmenting worker capabilities in the physical world through guidance has been a decades-long goal for AI agents. One key supporting capability is to understand what the worker is doing…
-
Do Syntax Heads Influence Model Behavior?
arXiv cs.CL
arXiv:2610.19231v1 Announce Type: new Abstract: Transformer language models develop attention heads that specialize in syntactic relations, for example, attending from verbs to their subjects. Do these heads play a causal role in the…
-
Delta Tokenizer: Reference-Identical Tokenization of Session Deltas and Segment-Parallelism for Long Prompts on Multicore CPUs
arXiv cs.CL
arXiv:2610.19801v1 Announce Type: new Abstract: An agent that drives an LLM sends the model the same conversation again and again, and each turn adds only a little text at the end. Inference engines reuse the KV cache of the shared prefix…
-
Does Machine Translation Retain Media Bias?
arXiv cs.CL
arXiv:2610.19945v1 Announce Type: new Abstract: Within the most recent two decades, machine translation (MT) has been advanced from a research area to a mature, deployed technology for prose text and most language pairs. However, language…
-
A Strong Harness Narrows the Model Gap: A Comparative Evaluation for Legislative Document-to-XML Conversion
arXiv cs.CL
arXiv:2610.19999v1 Announce Type: new Abstract: Legal-content reproduction requires matching the source exactly in section hierarchy, verbatim text, citations, and amendment spans; a dropped citation or mislabeled amendment carries into…
-
MCPDisCo: Seesaw-Mitigated Tuning for MCP-Native Agents
arXiv cs.AI
arXiv:2610.14956v1 Announce Type: new Abstract: The Model Context Protocol provides a shared interaction substrate for the agentic web, allowing heterogeneous services to expose capabilities through a standardized interface. Learning to…
-
Equilibrium Forcing: Adaptive Video Generation Without Noise Conditioning
arXiv cs.LG
arXiv:2608.14706v2 Announce Type: replace-cross Abstract: Standard autoregressive video generation algorithms based on Diffusion and Flow Matching rely on rigid training objectives and static sampling schedules, limiting inference…
-
LYNX: Analytical Program Synthesis via Grounding and Closed-Loop Refinement
arXiv cs.AI
arXiv:2610.14777v1 Announce Type: new Abstract: LLM-powered data analytics aims to translate natural language instructions into executable analytical programs over underlying data. However, existing approaches, whether treating LLMs as…
-
Beyond Size: How Gradients Shape Pruning Decisions in Large Language Models
arXiv cs.LG
arXiv:2311.04902v3 Announce Type: replace-cross Abstract: Large Language Models (LLMs) with billions of parameters are prime targets for network pruning, which removes some model weights without hurting performance. Prior approaches such…
-
Spectral structure learning for clinical time series
arXiv cs.LG
arXiv:2502.11680v2 Announce Type: replace Abstract: We develop and evaluate a structure learning algorithm for clinical time series. Clinical time series are multivariate time series observed in multiple patients and irregularly sampled…
-
Hypothesis Selection via Weighted Medians
arXiv cs.LG
arXiv:2610.20375v1 Announce Type: cross Abstract: Given $n$ distributions (hypotheses) $\mathcal H=\{H_1,\ldots,H_n\}$ and samples from an unknown target distribution $P$, proper hypothesis selection seeks a distribution whose total…
-
Policy-Verified Self-Training: Learning Agent Compliance from Verifiable Policy Rewards
arXiv cs.LG
arXiv:2610.19372v1 Announce Type: new Abstract: Runtime policy verifiers can prevent agents from executing disallowed actions, but their decisions are typically used only for enforcement and discarded afterward. We ask whether these…
-
Contrastive uncertainty learning for robust multi-structure segmentation in lower-pelvic MR
arXiv cs.LG
arXiv:2610.20251v1 Announce Type: cross Abstract: Boundary delineation of anatomical structures in prostate MRI is often compromised by regional ambiguity. Automated segmentation methods relying on global overlap loss functions tend to be…
-
Cross-Model In-Context Neurofeedback Reveals (Limited) Activation Control in LLMs
arXiv cs.LG
arXiv:2610.20179v1 Announce Type: cross Abstract: Probes on activation vectors are widely used to detect misbehavior in language models, but their success relies on the assumption that models cannot control what their activations encode…
-
EmoXpt: Analyzing Emotional Variances in Human Comments and LLM-Generated Responses
arXiv cs.AI
arXiv:2501.06597v1 Announce Type: cross Abstract: The widespread adoption of generative AI has generated diverse opinions, with individuals expressing both support and criticism of its applications. This study investigates the emotional…
-
MARL-BattleGrounds: A Novel JAX-native Benchmark for Heterogeneous and Competitive Multi-agent RL
arXiv cs.AI
arXiv:2610.14883v1 Announce Type: new Abstract: We present MARL-BattleGrounds (MARL-BGs), an easy-to-use JAX-native benchmark for heterogeneous and competitive multi-agent reinforcement learning (MARL), designed to be sample-efficient…
-
Hierarchical Residual Policy Learning for Real-World Mobile Manipulation with Sparse Human Guidance
arXiv cs.LG
arXiv:2610.20607v1 Announce Type: cross Abstract: Mobile Manipulation (MoMa) is a cornerstone of autonomous robotic systems, yet real-world fine-tuning remains a formidable challenge because of high-dimensional action spaces and intricate…
-
Toward System One Clinical Intelligence: Parameter-Efficient Fine-Tuning and Counterfactual Explainability of Decision Foundation Models
arXiv cs.AI
arXiv:2610.15029v1 Announce Type: cross Abstract: Automated disease classification from patient-reported symptoms requires balancing the speed of lightweight discriminative models against the contextual understanding and latency costs of…
-
Analyzing and Mitigating Poisoning Attacks on Flow-Based Intrusion Detection
arXiv cs.LG
arXiv:2610.19766v1 Announce Type: cross Abstract: Although Artificial Intelligence (AI)-based Intrusion Detection Systems (IDSs) are increasingly deployed in operational networks, their dependence on large-scale training data introduces a…
-
Articulation Upper Bounds for Language Models
arXiv cs.LG
arXiv:2610.19721v1 Announce Type: cross Abstract: Language captures only a subset of the perceptions, intuitions and intents that form meaning between speakers. What, then, are the limits of language? We take steps towards measuring this…
-
ANCHOR-RE: An Agentic Biomedical Relation Extraction Framework with Retrieval Augmentation and Error-Driven Verification
arXiv cs.CL
arXiv:2608.03154v2 Announce Type: replace Abstract: Biomedical relation extraction (BioRE) supports knowledge graph construction and hypothesis generation. In real-world settings, most candidate entity pairs are unrelated. Symbolic systems…
-
Physical Layer Anomaly Detection via Adaptive Gaussian Mixture Model
arXiv cs.LG
arXiv:2610.18946v1 Announce Type: cross Abstract: Physical-layer spoofing attacks pose a critical threat to wireless localization systems by manipulating channel state information (CSI). Detecting such attacks becomes particularly…
-
Distinguishing Process Change, Instrument Fault and Data-Path Corruption in Streaming Telemetry: A Reference Implementation and Benchmark
arXiv cs.LG
arXiv:2610.18904v1 Announce Type: cross Abstract: A reading departs from what a monitoring system expects for one of three reasons: the physical process changed, the instrument is failing, or the value was corrupted between the instrument…
-
When LLM Reward Design Fails: Diagnostic-Driven Refinement for Sparse Structured RL
arXiv cs.AI
arXiv:2605.28918v2 Announce Type: replace-cross Abstract: For sparse, structured reinforcement-learning tasks with semantic reward-function interfaces, LLM-generated reward shaping is better framed as debugging than one-shot generation. We…
-
From Conceptual EneMonIE to an Executable Backend: A Modular Architecture for Energy Information Extractionand Grounded Question Answering
arXiv cs.AI
arXiv:2610.17243v1 Announce Type: cross Abstract: The original EneMonIE proposal describes a Web-based semantic system for continuously monitoring energy trends through targeted information extraction from heterogeneous online media. This…
-
TacitVLA: Reasoning Vision-Language-Action Models Without Reasoning Latency
arXiv cs.AI
arXiv:2610.17540v1 Announce Type: new Abstract: Textual reasoning before acting has improved vision-language-action (VLA) models, but at the cost of latency: every action conditioned on the reasoning text must wait for autoregressive…
-
Beyond Expert Rankings: Contrast Calibration for Mixed-Precision MoE Quantization
arXiv cs.LG
arXiv:2610.20353v1 Announce Type: new Abstract: Mixture-of-experts (MoE) language models activate only a few experts per token, yet all expert weights must be stored. Mixed-precision quantization reduces this storage cost by assigning…
-
REA: From Correct LLM Judgments to Safe Decision Delivery
arXiv cs.AI
arXiv:2610.17473v1 Announce Type: new Abstract: Generative safety guards must classify requests correctly and return outputs that applications can parse. Reasoning guards can improve classification accuracy while producing outputs that…
-
When More Documents Hurt RAG: Measuring and Reducing Vector Search Dilution with Domain-Scoped Retrieval
arXiv cs.CL
arXiv:2606.11350v2 Announce Type: replace Abstract: Retrieval-augmented generation (RAG) degrades when applied to large, heterogeneous document collections. In RAG systems, dense similarity becomes less discriminative, causing top-$k$…