Research

Latest Research

A running feed of newly published AI research papers, pulled daily from arXiv and filtered by hand before it appears here.

Papers

50 items
Keyword
01Human-Centred, Risk-Constrained Evaluation of Financial Document Recognition: Evidence, Critical Errors, and Selective AutomationarXiv cs.AIarXiv:2611.02775v2 Announce Type: replace-cross Abstract: Financial document recognition is often assessed using character and word error rates, although these metrics do not express whether a system preserves the financial facts on which people and organisations act. This paper develops a human-centred, risk-constrained framework for evaluating financial document recognition as evidence recovery and decision support. It defines typed financial facts linked to visible source regions and separat06 Nov 2026↗02Multi-Estimate Elicitation Does Not Uniformly Debias Language Model JudgmentsarXiv cs.AIarXiv:2611.00597v2 Announce Type: replace-cross Abstract: Does eliciting five weighted estimates within a single response consistently improve numerical judgments? We compare direct and multi-estimate elicitation across anchoring, hindsight, base-rate, conjunction, and status-quo/default tasks in six frozen configurations. Each configuration contributes 618 admitted development units and 3,702 admitted formal units; the six-configuration formal cohort contains 22,212 units, including 21 missing06 Nov 2026↗03Revisiting Generative vs. Dense Retrieval: A Static Value Prior Reverses the GaparXiv cs.AIarXiv:2610.19065v2 Announce Type: replace-cross Abstract: Generative retrieval (GR) has emerged as a strong candidate generator for sequential recommendation as language models and semantic-item identifiers have improved. However, GR is not a uniformly stronger replacement of dense retrieval (DR) in online retail candidate generation. We examine their differences on two Amazon Toys datasets under different metrics. GR leads on best relevant item value among users both methods cover, while DR le06 Nov 2026↗04Autonomous Driving Research Requires a Community-Driven Data ParadigmarXiv cs.AIarXiv:2610.08825v2 Announce Type: replace-cross Abstract: Autonomous driving has made remarkable progress, with recent AI advances enabling commercial deployments that are reshaping urban mobility. Yet the field remains far from its universal social promise: autonomous systems that can operate robustly anywhere, anytime, for anyone. We posit that this gap is not merely a modeling problem, but a problem of the prevailing data paradigm. Current research relies heavily on a few benchmark datasets 06 Nov 2026↗05AnchorRep: Defending LLMs Against Cross-Model Adversarial Transfer via Representation RepulsionarXiv cs.AIarXiv:2609.32602v2 Announce Type: replace-cross Abstract: Adversarial attacks optimized on a single open-weight LLM can transfer to and jailbreak architecturally different models, allowing an attacker with white-box access to one model to compromise independently deployed systems. This creates a shared vulnerability across models, yet existing defenses are not designed for this cross-model threat. We find that cross-model transfer aligns with shared internal representation geometry, making it a06 Nov 2026↗06A Cyber Range Evaluation of Inductive Autonomous Network Incident Response AgentsarXiv cs.AIarXiv:2609.16541v2 Announce Type: replace-cross Abstract: We have applied a novel combination of classic relational database modeling and modern reinforcement learning to automated intrusion response to enable inductive decision agents. Unlike previous work in the area, our approach allows agents to handle variable quantities of network assets without retraining. We test the operational performance of our approach in a virtualized computer network with variable topologies, red-team automation a06 Nov 2026↗07AI Slop and Hallucinations in Vulnerability Assessment: A Survey on Reasoning Failures and Trustworthy MitigationarXiv cs.AIarXiv:2608.25667v2 Announce Type: replace-cross Abstract: The integration of Large Language Models (LLMs) into cybersecurity has transformed vulnerability assessment, but it has also produced a trustworthiness crisis driven by the unchecked proliferation of "AI slop." These artifacts, hallucinated vulnerabilities, plausible but incorrect patches, and semantically repackaged bug reports, impose a cognitive burden on human triage pipelines that mirrors a denial-of-service attack. This paper surve06 Nov 2026↗08Parameter Exploration for RLVR via Variational LearningarXiv cs.AIarXiv:2608.09805v2 Announce Type: replace-cross Abstract: Exploration has been a focus of reinforcement learning research for a long time. Recently, there has been growing evidence that it is also an important ingredient in LLM reinforcement learning recipes that can significantly impact downstream performance. Many existing methods control exploration in the action-space, for example, using temperature scaling. However, these methods cannot reorder tokens but only influence the variance in the06 Nov 2026↗09Eigenius: A Typed Knowledge-Graph DBMS with Epistemic Stratification and Institution-Mediated ReasoningarXiv cs.AIarXiv:2608.04457v3 Announce Type: replace-cross Abstract: As "AI Scientists" emerge to drive research via the Model Context Protocol (MCP), systems relying on ephemeral scripts will fail. The sheer scale of stateful, interconnected evidence requires a machine-walkable warranty grounded in a purpose-built database architecture. Eigenius is an open-source, typed knowledge-graph DBMS built on a single premise: answering the audit question ("what do you know, and what is your warranty?") requires a06 Nov 2026↗10Invariant Discovery for Networked SystemsarXiv cs.AIarXiv:2607.22944v2 Announce Type: replace-cross Abstract: Invariants, the relations expected to hold among measured signals of a network, underpin applications from verification to traffic generation, telemetry imputation, and input validation. Yet, writing them by hand demands rare expertise in both formal logic and networking. Testing all possible combinations of telemetry variables statistically is intractable at network scale, while existing automatic tools still require a human-authored gr06 Nov 2026↗11LatentGym: A Testbed For Cross-Task Experiential Learning With Controllable Latent StructurearXiv cs.AIarXiv:2606.15306v2 Announce Type: replace-cross Abstract: We envision continually learning agentic systems that become more useful over time: as they encounter sequences of related tasks, they should infer the hidden structure shared across those tasks and use it to improve future decisions. This cross-task experiential learning capability is pivotal in domains such as personalization and interactive assistance, but existing training/evaluation frameworks do not provide shared, controllable lat06 Nov 2026↗12The Standard Interpretable Model: Design interpretable machine learning methods from elementary symmetriesarXiv cs.AIarXiv:2606.12289v3 Announce Type: replace-cross Abstract: As Artificial Intelligence models grow in complexity, interpretability has become an indispensable tool for understanding, debugging, and controlling their computations. However, interpretability lacks general theories to deductively design interpretable methods. This gap between theories and methods results in a fragmented literature and inconsistent evaluation protocols. To fill this gap, we introduce the Standard Interpretable Model06 Nov 2026↗13Alignment Collapse Under KV Cache Quantization: Diagnosis and MitigationarXiv cs.AIarXiv:2606.09864v3 Announce Type: replace-cross Abstract: Key-value (KV) cache quantization is widely used to reduce LLM inference memory, yet existing evaluations focus only on perplexity and accuracy without assessing the safety impact. In this study, we explore alignment preservation under KV cache quantization. Across eleven instruction-tuned models (3.8B-72B) and five benchmarks (1,894 prompts), we find that low-bit quantization can silently destroy safety alignment: Mistral-7B loses 15.2%06 Nov 2026↗14Teach the Way It Thinks: Guiding LLM Post-training Data Engineering with Sparse AutoencodersarXiv cs.AIarXiv:2605.27354v2 Announce Type: replace-cross Abstract: Data engineering for large language model (LLM) reinforcement learning (RL) decides which samples to train on, when to present them, and which to batch together. Existing methods mostly rely on external feedback, such as pass rates or difficulty labels, which observes only the outcome on each sample. We argue that model internals offer richer and more model-intrinsic feedback that can guide all three decisions. We propose SAERL which org06 Nov 2026↗15JudgmentBench: Comparing Rubric and Preference Evaluation for Quality AssessmentarXiv cs.AIarXiv:2605.25240v4 Announce Type: replace-cross Abstract: Two methodologies dominate current practices of benchmarking: rubric-based scoring evaluates items against predefined criteria, whereas comparative judgment elicits pairwise preferences between outputs. Although both methodologies are widely used, the choice between them is rarely justified. We release JudgmentBench, a benchmark of 30 real-world legal tasks, paired with 1,539 rubric scores and 1,530 pairwise preference judgments collecte06 Nov 2026↗16Vector Policy Optimization: Training for Diversity Improves Test-Time SearcharXiv cs.AIarXiv:2605.22817v2 Announce Type: replace-cross Abstract: Language models must now generalize out of the box to novel environments and work inside inference-scaling search procedures, such as AlphaEvolve, that select rollouts with a variety of task-specific reward functions. Unfortunately, the standard paradigm of LLM post-training optimizes a pre-specified scalar reward, often leading current LLMs to produce low-entropy response distributions and thus to struggle at displaying the diversity th06 Nov 2026↗17An Assessment of Human vs. Model Uncertainty in Soft-Label Learning and CalibrationarXiv cs.AIarXiv:2605.18648v3 Announce Type: replace-cross Abstract: Central to human-aligned AI is understanding the benefits of human-elicited labels over synthetic alternatives. While human soft-labels improve calibration by capturing uncertainty, prior studies conflate these benefits with the implicit correction of mislabeled data (mode shifts), obscuring true effects of soft-labels. We present a controlled audit of soft-label learning across MNIST and a synthetic variant, re-annotating subsets to ext06 Nov 2026↗18Event-Grounded Sparse Autoencoders for Vision-Language-Action PoliciesarXiv cs.AIarXiv:2605.17204v2 Announce Type: replace-cross Abstract: Vision-Language-Action (VLA) policies translate language and visual inputs into robot actions, where their hidden representations directly shape closed-loop behavior. However, mechanistic interpretability tools from language and vision-language models do not transfer cleanly to VLAs: outputs are robot actions rather than human-readable tokens, and validating internal changes requires expensive closed-loop rollouts. We propose an event-gr06 Nov 2026↗19PanoWorld: Geometry-Consistent Panoramic Video World ModelingarXiv cs.AIarXiv:2605.15391v2 Announce Type: replace-cross Abstract: We present PanoWorld, a panoramic video world model that generates geometry-consistent 360$\degree$ video from a single image and a caption. Existing panoramic video methods optimize primarily for visual realism and do not explicitly constrain the underlying 3D scene state, producing outputs that appear plausible yet exhibit inconsistent depth, broken correspondences, and implausible motion across the spherical surface. We address this g06 Nov 2026↗20EcoGEO: Trajectory-Aware Evidence Ecosystems for Web-Enabled LLM Search AgentsarXiv cs.AIarXiv:2605.12887v3 Announce Type: replace-cross Abstract: Web-enabled LLM agents are changing how online information influences search outcomes. Existing Generative Engine Optimization (GEO) studies mainly focus on individual webpages. However, agentic web search is not a single-document setting: an agent may issue queries, crawl pages, follow links, reformulate searches, and synthesize evidence across multiple browsing steps. Influence therefore depends not only on page content, but also on ho06 Nov 2026↗