114
归档文章
6
活跃日期
8
新闻源
来源概览
arXiv cs.AI
66 articles
Latest: Nov 9, 2026
arXiv cs.CL
18 articles
Latest: Nov 9, 2026
arXiv cs.LG
15 articles
Latest: Nov 9, 2026
Hacker News
4 articles
Latest: Nov 9, 2026
The Decoder
4 articles
Latest: Nov 8, 2026
MIT Tech Review AI
2 articles
Latest: Nov 9, 2026
Towards AI
1 article
Latest: Nov 9, 2026
Wired AI
1 article
Latest: Nov 6, 2026
Nov 9, 2026
104Innovating with AI in the special chemicals industry
As companies look to accelerate innovation with artificial intelligence, cloud computing, and advanced modeling, the primary challenge is connecting new technologies to reliable data, scientific expertise, and the pract…
Hours of YouTube and 85 handtyped pages of notes on LLM Inference – Start here
Article URL: Comments URL: Points: 26 # Comments: 2
Claude Code Missed $800K in Contracts. Here’s How to Check Yours
An AlixPartners talk showed how messy contract data fooled Claude Code. In 510 real contracts, 18% auto-renew with a notice period, so the real deadline comes before the end date. Claude Code gave a consulting team a co…
ChatGPT Desktop forwards your SSH agent to remote machines
The agent running on a remote machine can authenticate using your local SSH credentials. That means it can push to your private repos on GitHub, for example. Proof: Comments URL: Points: 76 # Comments: 17
Is Opus 5.5 nerfed? Thinking patterns changed on November 3
Article URL: Comments URL: Points: 294 # Comments: 202
How Much Conversation History Does an AI Agent Need?
arXiv:2611.09662v1 Announce Type: cross Abstract: How does using less conversational history change reader-token use and benchmark scores? We examine this question on 1,100 BEAM questions. We held the retrieval ranking…
Internal Safety Collapse in LLMs and Coding Agents
arXiv:2603.23509v2 Announce Type: replace-cross Abstract: LLMs and coding agents can now tackle complex, long-horizon tasks through advances in reinforcement learning, but stronger task-solving capabilities do not neces…
Debate to Deceive: Poisoning LLM's Fact-Checking via Game-Theoretic Multi-Agent Debate
arXiv:2611.08756v1 Announce Type: new Abstract: Knowledge attacks on Large Language Models, such as manipulating fact-checking judgments, pose significant risks to AI reliability. Existing approaches -- adversarial suff…
Toward LLM Self-Play Under Data Scarcity: Effectiveness, Generalizability, and Trustworthiness
arXiv:2611.08908v1 Announce Type: new Abstract: The rapid development of large language models (LLMs) has been driven largely by scaling model size and human-curated training data, yet this paradigm is increasingly cons…
Decision Context: Persistent Behavioral Learning for AI Agents
arXiv:2611.08644v1 Announce Type: new Abstract: LLM agents do not learn from corrections across sessions. A user fixes the same mistake at the start of every conversation. Existing agent memory systems primarily handle…
Cancellation Is Not Enough: What Retries Leave Behind in an Overloaded LLM Engine
arXiv:2611.08684v1 Announce Type: new Abstract: Clients and proxies in front of an LLM engine retry calls that time out. After a transient overload, these retries can keep the engine from recovering even when calls arri…
Memory-Aware Multi-Agent Architecture with LLM-based Orchestration for Conversational Recommendation
arXiv:2611.08718v1 Announce Type: new Abstract: Conversational recommender systems often treat user sessions in isolation, limiting their ability to capture how user intent evolves over time. While large language models…
Solver-Admitted Metamorphic Testing of Learned Network Optimization Surrogates
arXiv:2611.08837v1 Announce Type: new Abstract: Equivalent encodings of an optimization problem can produce different learned predictions. Interpreting these differences requires accounting for numerical error and multi…
When F1 Is Not Enough: Auditing Codebooks for LLM-Assisted Annotation
arXiv:2611.08846v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly used for text annotation in social science, but standard performance metrics do not explain why errors occur. Low F1 may refl…
Mind the Divergence: Divergence-Guided Reliability-Aware Fusion and Adaptive LLM Prompting for Conversational Emotion Recognition
arXiv:2611.08860v1 Announce Type: new Abstract: Multimodal emotion recognition in conversations (MERC) integrates textual, acoustic, and visual cues to identify utterance-level emotions. However, existing approaches gen…
Analyzing Code Provenance in LLMs for Programming Education
arXiv:2611.09177v1 Announce Type: new Abstract: Large language models (LLMs) play an increasing role in programming education, challenging traditional assessment methods as they can generate fully functional code instan…
Large Language Model Consensus and Corpus-Adaptive Topic Modeling for Classifying and Mapping Patent Innovation
arXiv:2611.09194v1 Announce Type: new Abstract: Patent analytics increasingly relies on transformer representations and large language models (LLMs), yet linking conceptual classification, topic formation and interpreta…
RAG-Enhanced LLMs for Interactive Explainability in Clinical Decision Support Systems
arXiv:2611.09232v1 Announce Type: new Abstract: Artificial Intelligence (AI) is increasingly embedded in high-risks decision-making systems across domains such as healthcare, finance, and the judicial system. As relianc…
Nudging Decision Models: Do Digital Nudges Steer Fast, Generation-Free AI Agents?
arXiv:2611.09264v1 Announce Type: new Abstract: LLM-based agents increasingly automate business and consumer tasks but are slow and expensive. Decision models, a new class that chooses among options without generating t…
PHISHREV+: An ASP-Based Post-Hoc Defeasible Machine Learning Framework for Phishing Detection
arXiv:2611.09273v1 Announce Type: new Abstract: Phishing website detection remains a critical cybersecurity challenge. Although machine learning classifiers achieve high accuracy on benchmark datasets, they suffer from…
Definitions, Label Names, or Demonstrations? Auditing What Governs LLM Annotation
arXiv:2611.09457v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly used to annotate social-science texts, but high accuracy does not establish that they apply researcher-supplied category defi…
LEMMO: A Large Electromagnetic Model leveraging Pretrained LLMs for Million-Sample Signal Understanding
arXiv:2611.09476v1 Announce Type: new Abstract: Large language models have expanded beyond text to vision, audio, and other modalities, but understanding raw electromagnetic (EM) signals remains challenging. Wideband re…
Intent-Aware Caching for Efficient LLM Serving
arXiv:2611.09715v1 Announce Type: new Abstract: Modern LLM serving systems use prefix caching to accelerate multi-turn conversational workloads. They cache the key-value states of previous conversation prefixes and reus…
Knowing When to Intervene: Causal Evaluation of AI Agents
arXiv:2611.09792v1 Announce Type: new Abstract: Current evaluations of agentic AI systems primarily rely on benchmarks. However, the static nature of benchmarks makes them uninformative about the value of auditing inter…
Ensemble Methodology for Reference-Free Confidence Estimation in LLM-based Code Summarization
arXiv:2611.09846v1 Announce Type: new Abstract: This paper describes an approach and use case to estimate the correctness of LLM-generated COBOL summaries without access to human-authored references. The core intuition…
SAFTAC: Simulation-Augmented Fine-Tuning of Open-Source LLMs for Analog Circuit Design
arXiv:2611.08543v1 Announce Type: cross Abstract: Circuit design for analog integrated circuits (ICs) is challenging because circuit topology, device sizing, and circuit performance are tightly coupled. Existing large l…
CascadeRec: Deployed Adaptive LLM Compute Allocation for Multi-Tenant Content Recommendation
arXiv:2611.08566v1 Announce Type: cross Abstract: Large language models (LLMs) are attractive rerankers for recommendation. Running an expensive model on every request is costly, and for users with sparse interaction hi…
Patching the Blind Spots: Reversible LLM Rule Overlay for Frozen Fraud Detectors
arXiv:2611.08714v1 Announce Type: cross Abstract: In industrial tabular fraud classification, gradient-boosted trees remain state-of-the-art, but even a well-trained model has blind spots, and retraining the deployed mo…
How are viral trends different from educational and commercial uses of LLM image generation and editing? A privacy-preserving descriptive analysis
arXiv:2611.08749v1 Announce Type: cross Abstract: Gemini's Nano Banana models, first introduced in late 2025, led to a shift in the capabilities and use of image generation and editing. We utilize Urania, a Google-devel…
Guardrail Erosion in Domain Fine-Tuned Large Language Models: A Paired Audit of Knowledge-Centric Abstention in Biomedical LLMs
arXiv:2611.08780v1 Announce Type: cross Abstract: The number of domain fine-tuned models, particularly in the biomedical domain, is rapidly expanding. Fine-tuning promises high-performing models at a fraction of the cos…
MotionDecipher: LLM-Assisted Clinical Interpretation of Human Motion Signals for Efficient Gait Analysis
arXiv:2611.08795v1 Announce Type: cross Abstract: Clinical gait and motion analysis is widely used to guide targeted interventions for patients with neurological disorders, the outcome of which is typically a report wit…
Phantom Consensus: Measuring Social Reasoning in Multi-Agent LLM Decision-Making
arXiv:2611.08971v1 Announce Type: cross Abstract: Multi-agent LLM systems increasingly aggregate judgments across model instances, but reasoning about other agents can itself introduce failure. We study a setting in whi…
Same Facts, Different Remedy: Linguistic Profiling and Arbitrariness in LLM and System One Decisions
arXiv:2611.08995v1 Announce Type: cross Abstract: Large language models are increasingly used to resolve consumer complaints, and the way complainants write varies with age, education and socioeconomic status (SES). We…
Seeing vs. Counting: Complexity Twins Show Where Multimodal LLMs Fail on Diagrams
arXiv:2611.08998v1 Announce Type: cross Abstract: When a multimodal LLM miscounts in a diagram, does it fail at seeing or at counting? Existing visual graph and puzzle benchmarks vary visual density and algorithmic hard…
When Routine Operations Break Architectural Pattern Relationships in LLM Agent Harnesses
arXiv:2611.09088v1 Announce Type: cross Abstract: LLM agent harnesses distribute tool mediation, state management and safety guardrails across native services, independently written plugins and configuration. Routine op…
On the Limits of Execution Traces for Repairing Assertion Failures in LLM-Generated Code
arXiv:2611.09212v1 Announce Type: cross Abstract: Iterative repair returns a failed program's error to a large language model (LLM) and asks it to fix it. After iterative repair, the remaining failures are mostly assert…
Can LLMs Debug Robots by Editing Their Weights?
arXiv:2611.09376v1 Announce Type: cross Abstract: A language model can repair a robot policy's lexical instruction grounding by editing just 65 token-embedding rows. We bring weight-space programming to robot foundation…
Query-Guided RoPE Modulation for Lost-in-the-Middle Mitigation in Long-Context LLMs
arXiv:2611.09554v1 Announce Type: cross Abstract: Long-context language models often neglect information in the middle of the input, a phenomenon known as "lost-in-the-middle". Part of this gap may stem from the long-ra…
Verifier Noise Is a Temperature: Testing Tilt Predictions in Finite-Budget LLM-Guided Search
arXiv:2611.09699v1 Announce Type: cross Abstract: LLM-guided search loops propose a modified candidate, score it, and keep it if it scores better. When both candidates are re-scored at every comparison, as in the miniba…
MemoryDojo: Benchmarking Poisoning Through LLM Agent Memory
arXiv:2611.09721v1 Announce Type: cross Abstract: LLM agents benefit from using memory systems to improve performance on long-horizon, multi-session tasks. However, memory introduces a critical attack surface. Malicious…
Language as Trait, Politeness as Strategy: Controlled Tests of Linguistic Effects in LLM Social-Deduction Agents
arXiv:2611.09768v1 Announce Type: cross Abstract: Large language model (LLM) agents coordinate through conversation, and the same message can be phrased in simpler or more complex language, bluntly or politely. We test…
Counterfactual Distillation: Internalizing Reflective Experience into LLM Agent
arXiv:2603.16843v2 Announce Type: replace Abstract: Large language models increasingly act as autonomous agents in long-horizon interactions, learning from final outcomes and environmental feedback. However, optimizing…
ClawTrace: Cost-Aware Tracing for LLM Agent Skill Distillation
arXiv:2604.23853v3 Announce Type: replace Abstract: Skill-distillation pipelines learn reusable rules from LLM agent trajectories, but they lack a key signal: how much each step costs. Without per-step cost, a pipeline…
From Senses to Decisions: The Information Flow of Auditory and Visual Perception in Multimodal LLMs
arXiv:2606.10147v3 Announce Type: replace Abstract: Multimodal Large Language Models (MLLMs) can listen and see, but how do audio and visual signals actually travel through the network to shape an answer? Despite their…
GABench: A Comprehensive Benchmark for Evaluating LLM Agents on Graph Analysis Tasks
arXiv:2608.01684v2 Announce Type: replace Abstract: Large language model (LLM) agents are increasingly capable of planning, using tools, and interacting with external environments. They are typically supported by harnes…
TailSieve: Partial-Rollout-Guided Tail Routing for LLM Rollouts
arXiv:2608.22788v3 Announce Type: replace Abstract: Large-scale rollouts have become a core component of modern LLM systems, spanning reinforcement learning (RL) post-training, on-policy distillation, and sampling-heavy…
MoMHa: Multi-Objective Optimization of LLM Harnesses over Accuracy, Safety, and Tokens
arXiv:2609.30967v2 Announce Type: replace Abstract: Most work on improving large language models treats accuracy as the sole objective. We argue that the harness, the Python code surrounding the model that constructs pr…
Programs-of-Layers in LLMs through the Lens of Cortical Areas
arXiv:2609.31360v2 Announce Type: replace Abstract: Inference in LLMs is conventionally a fixed-depth, fixed-order forward pass through every layer, regardless of how difficult the input is. The human brain does not wor…
AuthGateBench: Benchmarking LLMs as Authorization Gate
arXiv:2610.16426v2 Announce Type: replace Abstract: Automated approval systems let coding agents execute high-risk actions with minimal human oversight by delegating the authorization decision to a classifier, typically…
SteadyThought: Mitigating LLM Under-Thinking via Thought-Level Preference Optimization
arXiv:2610.25045v2 Announce Type: replace Abstract: Flexible switching between reasoning trajectories (i.e., thought switching) has significantly enhanced the reasoning capabilities of Large Reasoning Models (LRMs). How…
A Definition of Catastrophic Performance
arXiv:2611.06219v2 Announce Type: replace Abstract: A central challenge in AI is specifying objectives that align with human preferences. While existing theoretical results elucidate how misspecification impacts perform…
End-to-End On-Device Quantization-Aware Training for LLMs at Inference Cost
arXiv:2509.00031v3 Announce Type: replace-cross Abstract: Quantization reduces the deployment cost of large language models (LLMs). However, post-training quantization (PTQ) cannot fine-tune model parameters and often d…
Knowledge Localization in Mixture-of-Experts LLMs Using Cross-Lingual Inconsistency
arXiv:2603.17102v2 Announce Type: replace-cross Abstract: Modern LLMs continue to exhibit significant variance in behavior across languages, such as being able to recall factual information in some languages but not oth…
Fast and Accurate Probing of In-Training LLMs' Downstream Performances
arXiv:2604.01025v2 Announce Type: replace-cross Abstract: The paradigm of scaling Large Language Models (LLMs) in both parameter size and test time has pushed the boundaries of AI capabilities, but at the cost of making…
Models Recall What They Violate: Constraint Adherence in Multi-Turn LLM Ideation
arXiv:2604.28031v3 Announce Type: replace-cross Abstract: When researchers refine ideas with large language models over several turns, do the models preserve the original brief's constraints? We introduce DriftBench, a…
Local Intrinsic Dimension Unveils Hallucinations in Diffusion Models
arXiv:2605.05026v2 Announce Type: replace-cross Abstract: Diffusion models are prone to generating structural hallucinations - samples that match the statistical properties of the training data yet defy underlying struc…
Reformulating KV Cache Eviction Problem for Long-Context LLM Inference
arXiv:2605.07234v2 Announce Type: replace-cross Abstract: Large language models (LLMs) support long-context inference but suffer from substantial memory and runtime overhead due to Key-Value (KV) Cache growth. Existing…
SCDBench: A Benchmark for LLM-Based Smart Contract Decompilers
arXiv:2605.29059v2 Announce Type: replace-cross Abstract: Smart contract decompilation aims to recover high-level source code from bytecode, but evaluating decompilers remains difficult because existing studies use narr…
Exposing the Illusion of Erasure in Knowledge Editing for LLMs
arXiv:2606.23276v3 Announce Type: replace-cross Abstract: Knowledge Editing (KE) has emerged as a frontier for updating specific facts in LLMs without costly retraining, but its reliability and underlying mechanisms rem…
MCP Server Architecture Patterns for LLM-Integrated Applications
arXiv:2606.30317v2 Announce Type: replace-cross Abstract: The Model Context Protocol (MCP), introduced by Anthropic in November 2024, defines a standardized interface for connecting large language models (LLMs) to exter…
Decomposing Yes-No Bias in LLM Moral Judgment: Stance, Position, Word, and Verdict
arXiv:2607.05552v2 Announce Type: replace-cross Abstract: Large language models (LLMs) increasingly issue binary verdicts - safety gates, judge decisions, value probes - and the answer is read as the model's judgment. A…
Beyond Bayesian Nash: Learning Minimax-Regret Equilibria for Adversarial Team Games under Asymmetric Information
arXiv:2607.09993v2 Announce Type: replace-cross Abstract: Adversarial team games (ATGs) with asymmetric information, such as adversarial path-finding, goal search, and reachability games on graphs, require strategies th…
Broken Gates: Re-evaluating Web Bot Defenses in the Age of LLM Agents
arXiv:2607.18659v2 Announce Type: replace-cross Abstract: The rapid emergence of web automation across commercial crawlers, solver services, and LLM-based browser agents has created new challenges for defenses that must…
Learning to Persuade Exposes How Easily LLMs Abandon Correct Beliefs
arXiv:2608.11624v2 Announce Type: replace-cross Abstract: Persuasion is a core dynamic of natural language communication, shaping how large language models (LLMs) update beliefs, resolve disagreements, and reach decisio…
BiG-SURE - Bipartite Graph for Semantic Uncertainty and Reliability Estimation of LLMs
arXiv:2608.30646v3 Announce Type: replace-cross Abstract: Reliable uncertainty estimation is a crucial requirement for deploying large language models (LLMs) and vision-language models (VLMs) in safety-critical settings…
Anatomy of LLM Sycophancy: What a Flip Rate Hides
arXiv:2610.06522v2 Announce Type: replace-cross Abstract: A model under pushback can correct itself, capitulate, or hold, and one flip rate counts a correction and a capitulation alike. SycoLens, a modular replay protoc…
What Makes a Question Curious? Measuring Model Curiosity in Human-LLM Interactions
arXiv:2610.33950v2 Announce Type: replace-cross Abstract: Large language models increasingly act as partners in everyday communication. In these interactions, useful assistance may require not only answering, but also a…
Prompt Lottery in LLM-as-a-Judge: Measuring and Mitigating Sensitivity to Syntactic Rephrasing
arXiv:2610.38076v2 Announce Type: replace-cross Abstract: LLM-as-a-Judge has become a standard for evaluating large language models, yet most approaches implicitly assume that a single judge prompt functions as a stable…
LLMs as Judges and Examinees for Medical Item Difficulty Prediction
arXiv:2610.38402v2 Announce Type: replace-cross Abstract: Predicting medical item difficulty from item text alone remains challenging. Prior work has largely relied on linguistic features or text representations derived…
When Disclosure Is Compulsory: Evaluating Ride-Hailing Rebalancing Policies with Self-Interested LLM Driver Agents
arXiv:2610.41132v3 Announce Type: replace-cross Abstract: The EU Platform Work Directive, to be transposed by December 2026, requires ride-hailing platforms to state in plain language the main parameters of their automa…
CuriousTogether: Implications of Curiosity and Cultural Grounding in Multi-Agent LLM Interactions
arXiv:2611.01301v2 Announce Type: replace-cross Abstract: Large Language Models (LLMs) are increasingly deployed in multi-agent settings, exhibiting emergent collective behaviors, yet the role of curiosity as a driver o…
Richness Is Not Evenness: Ordered Renyi Crossings under Self-Training
arXiv:2611.08892v1 Announce Type: new Abstract: Recursive self-training of language models, a setting studied in model collapse, can reduce vocabulary while making frequent words, mostly function words, less dominant. W…
Using an LLM to Tackle Difficult Cases in NER for Nepali Tweets
arXiv:2611.08967v1 Announce Type: new Abstract: Named entity recognition (NER) is challenging for social media data in a low-resource language. LLMs often lag behind pretrained language models (PLMs) in NER performance,…
Beyond Truncation? Entailment-Licensed Elision and the Limits of Semantic Preservation in LLM Decoding
arXiv:2611.08975v1 Announce Type: new Abstract: Instruction-tuned large language models (LLMs) answer at greater length than a request needs. We ask whether a training-free controller inside the decoding loop can remove…
KLEAR: Knowledge-Matched LLM Evaluation in Korean across Application Requirements via Controlled Benchmark Transformation
arXiv:2611.09223v1 Announce Type: new Abstract: Existing Korean benchmarks for Large Language Models (LLMs) evaluate performance across diverse knowledge domains, but their accuracy scores cannot distinguish whether inc…
On the Effectiveness and Limitations of Open-Weight LLMs for Low-Resource Offensive Language Detection in Danish
arXiv:2611.09506v1 Announce Type: new Abstract: We study the effectiveness and robustness of open-weight large language models (LLMs) for Danish offensive language detection, an underexplored low-resource task. We evalu…
Benchmarking Speaker Embedders in the Sub-Second Regime for Real-Time Voice AI Agents
arXiv:2611.09516v1 Announce Type: new Abstract: Real-time Voice AI agents must separate the active user from intelligible background speech that survives conventional noise suppression. Background sources include televi…
Are These Programmers Ripe Enough to Eat? Evaluating How LLMs Handle Lexical Knowledge Conflicts in Translation
arXiv:2611.09785v1 Announce Type: new Abstract: External lexical resources can help large language models (LLMs) translate texts in low-resource languages and specialized domains, but these resources may conflict with m…
On Seeding Watermarks to Detect Verbatim LLM Copy-Paste Responses
arXiv:2605.16336v2 Announce Type: cross Abstract: Large language models (LLMs) have made fluent essay writing, code drafting, and quiz answering instantly available to students at every level, from secondary school thro…
Same Question, Different Answers: Auditing Persona-Dependent Variation in Six LLMs on Smoking-Cessation Advice
arXiv:2610.42039v1 Announce Type: cross Abstract: Hundreds of millions of people now ask large language model (LLM) chatbots for health advice, which raises a question that is simple to pose and hard to answer: does eve…
Copies Are Not Confirmations: How News Duplication Fools LLM Judges and Continual Learners
arXiv:2611.09203v1 Announce Type: cross Abstract: Many news reports of a claim are copies of one source: wire stories, content shared within a media group, syndicated features and press releases. We ask whether LLM judg…
Can LLMs Take Retrieved Information with a Grain of Salt?
arXiv:2605.06919v2 Announce Type: replace Abstract: Large language models have demonstrated impressive retrieval-augmented capabilities. However, a crucial area remains underexplored: their ability to appropriately adap…
PersonalAI 2.0: Enhancing knowledge graph traversal/retrieval with planning mechanism for Personalized LLM Agents
arXiv:2605.13481v5 Announce Type: replace Abstract: We introduce PersonalAI 2.0 (PAI-2), a novel framework designed to enhance LLM-based systems through integration of external knowledge graphs (KGs). The proposed appro…
Wind Turbine Maintenance Log Labelling Framework: LLM-Driven Data Correction and Enrichment via Semantic Extraction of Reliability Intelligence
arXiv:2605.31281v3 Announce Type: replace Abstract: Inconsistent codes and unstructured text limit the use of wind turbine maintenance records in reliability analysis. A topology-aware large language model (LLM) workflo…
Consolidating Rewarded Perturbations for LLM Post-Training
arXiv:2605.31494v2 Announce Type: replace Abstract: Post-training of language models is commonly framed as a sample-score-update loop implemented by gradient descent. A recent line of work, exemplified by RandOpt, reloc…
Sampling Reveals Style: Unsupervised, Training-Free Discovery of Prompt-Conditional Stylistic Axes in LLM Activations
arXiv:2609.19150v3 Announce Type: replace Abstract: Large language models (LLMs) encode rich stylistic structure in their hidden activations, but discovering which stylistic dimensions are salient for a given prompt typ…
Measure Before You Train: Typed Decision Models vs. a Classical Baseline vs. a Generative LLM for Routing a Personal Multilingual Inbox
arXiv:2610.32955v3 Announce Type: replace Abstract: Personal email agents decide, for every message, its type and urgency and whether to draft a reply, create a task or ignore it. In production these decisions are deleg…
TraceSpec: Understanding and Improving Speculative Decoding for LLM Agents
arXiv:2610.41177v2 Announce Type: replace Abstract: LLM agents execute tasks through repeated model calls over evolving histories of reasoning, actions, and tool feedback. Aggregate speculative-decoding metrics merge th…
Whispers from the Neighbors: Rethinking Caption Consensus as Inference-Time Evidence for Vision-Language Retrieval
arXiv:2610.33798v2 Announce Type: replace-cross Abstract: An image is rarely described only once. Retrieval benchmarks and many collections attach several captions to each picture, yet dual encoders still judge every im…
Interpretable Multifidelity Learning with Kolmogorov-Arnold Networks for Spatiotemporal Nuclear Reactor Flow Modeling
arXiv:2611.08512v1 Announce Type: new Abstract: High-fidelity computational fluid dynamics (CFD) simulations of nuclear reactor systems are computationally expensive, motivating machine learning methods that reduce cost…
Reconstructing LLM Lineage Trees via Fine-Tuning Path Inference
arXiv:2611.08866v1 Announce Type: new Abstract: Fine-tuned (FT) derivatives of large language models (LLMs) are now released in large numbers, yet their provenance relies on self-reports and cannot be verified. Misattri…
One Run Is Not Enough: A Monte Carlo Protocol for Evaluating LLM Reliability on Data Science Tasks
arXiv:2611.09729v1 Announce Type: new Abstract: Large language models (LLMs) have shown promise for automating data science workflows. Benchmarks such as DataSciBench (2026) and StatQA (2024) have advanced the evaluatio…
Reweighting for Transfer Classification: Exact Risk Conversion and Minimax Adaptation
arXiv:2611.08876v1 Announce Type: cross Abstract: A source can make a training mixture easier to learn while making its accuracy less valuable for the target. We study how the training problem a source creates and the f…
Authorization Drift in LLM Agents: Evaluating Boundary Compliance in Multi-Step Tool Use
arXiv:2611.09113v1 Announce Type: cross Abstract: LLM agents can exceed their authorization boundaries while pursuing user goals without an adversary. We formalize this phenomenon as authorization drift across data, too…
ReplayAudit: Auditing Outsourced LLM Fine-Tuning via Sealed Records and Selective Replay
arXiv:2611.09607v1 Announce Type: cross Abstract: Auditing outsourced LLM fine-tuning requires checking an untrusted provider's training while controlling verification cost. Uniform sampling ignores training-time signal…
Minerva: Reinforcement Learning with Verifiable Rewards for Cyber Threat Intelligence LLMs
arXiv:2602.00513v4 Announce Type: replace Abstract: Cyber threat intelligence (CTI) analysts routinely convert noisy, unstructured security artifacts into standardized, automation-ready representations. Although large l…
The Silent Hyperparameter: Quantifying the Impact of Inference Backends on LLM Reproducibility
arXiv:2605.19537v3 Announce Type: replace Abstract: Progress in LLMs is increasingly measured through standardized benchmarks, where state-of-the-art improvements are often separated by fractions of a percentage point.…
Dominant-Layer ZO: A Single Layer Dominates Zeroth-Order Fine-Tuning of LLMs
arXiv:2606.05516v2 Announce Type: replace Abstract: Zeroth-order (ZO) optimization enables memory-efficient fine-tuning of large language models (LLMs) using only forward passes, but it remains unclear how useful adapta…
The Sparsity Whisperer
arXiv:2608.06630v2 Announce Type: replace Abstract: Pruning reduces the inference cost of large language models, but existing criteria primarily preserve large activations or reconstruct layer outputs. We argue that thi…
Reward Budgeting Reduces Premature Convergence in Reinforcement Learning for LLM Reasoning
arXiv:2611.05942v3 Announce Type: replace Abstract: RL for LLM reasoning often plateaus early as its policy sharpens. Policy entropy is studied as both a diagnostic and a control signal, but increasing it alone does not…
Linear Bandits on Ellipsoids: Minimax Optimal Algorithms
arXiv:2502.17175v2 Announce Type: replace-cross Abstract: We consider linear stochastic bandits where the set of actions is an ellipsoid. We provide the first known minimax optimal algorithm for this problem. We first d…
Backdoor Channels Hidden in Latent Space: Extending Cryptographic Undetectability to Modern Neural Networks
arXiv:2605.13214v4 Announce Type: replace-cross Abstract: Recent cryptographic results establish that neural networks can be backdoored such that no efficient algorithm can distinguish them from a clean model. These gua…
Sharp Minimax Regret for Infinite-Memory Logistic Prediction
arXiv:2608.26515v3 Announce Type: replace-cross Abstract: We study sequential prediction of binary marks whose log-odds are a linear function of all past inputs. The inputs are independent uniform signs. The coefficient…
Same Request, Different Answer: Quantization Amplifies Cache-Induced Divergence in LLM Serving
arXiv:2609.04748v3 Announce Type: replace-cross Abstract: Major open-source serving stacks reuse the key-value tensors of a shared prefix by default. We measure what this prefix reuse costs in reproducibility. Holding m…
Nov 8, 2026
3Andrej Karpathy built a curated video platform in a weekend using Claude and 20 years of Hacker News
OpenAI co-founder Andrej Karpathy used Claude's Opus 5.5 to filter 43 million Hacker News comments into 6,454 curated videos. The result beats YouTube's own recommendations, he says. The article Andrej Karpathy built a…
The inventor of the bitter lesson has a bitter lesson for LLM scaling enthusiasts
Turing Award winner Richard Sutton reaffirms that raw compute alone isn't enough for intelligence and that fundamentally new learning methods are needed. Ex-OpenAI researcher Tworek agrees, saying blind transformer scal…
Llama-modchip: 284B DeepSeek on 64GB RAM, no GPU
Article URL: Comments URL: Points: 440 # Comments: 196
Nov 7, 2026
2Thinking Machines' Inkling 2 Small drops 316 F-bombs across benchmark runs
Thinking Machines' AI model Inkling 2 Small swears during its internal reasoning, 316 times across 2,655 benchmark runs, according to an analysis by Vals AI. The profanity typically shows up late in long runs that fail…
The gap between open and closed AI models in cyber capabilities keeps shrinking
In the Vals AI benchmark, OpenAI's GPT-6 Sol Daybreak hits 80.8 percent of attack objectives in simulated corporate networks. Worse, the open-weight model DeepSeek V4.1 Flash reaches 72.8 percent at a tenth of the cost.…
Nov 5, 2026
2Meta says US users can now video chat with their Muse AI avatar
Meta's AI agent can now take calls as a customizable animated avatar, ahead of the launch of a Tamagotchi-like wearable designed to bring Muse to life.
Gemini Omni 1.2 Flash: Fast and cost-effective with longer clips
Gemini Omni 1.2 Flash brings native 15-second video generation and editing, stronger consistency, and 30% lower pricing.
Nov 3, 2026
2Qwen 3.8 27B on my laptop can interpret model weights, too
Yesterday's hot paper was Sparks of Weight-Space Programming: LLMs can natively read & write neural network weights by Rikiya Takehi, Jayden Teoh, Alex Zhang, and Omar Khattab. The paper (OK, the authors classify is as…
How “reverse game theory” could keep AI agents in check
A new approach known as “reverse game theory” aims to control powerful AI models by designing systems that incentivize them to stick to the rules. As AI models become more capable, safety researchers are scrambling to e…