Predictability

Follow this topic Ask about this topic

Momentum

2 papers in the last four weeks, down 33% on the four weeks before. 0.0% of all new papers.

Jul 6Week of Sep 21

Latest papers 29

All topics

Jailbreak Context Lingers: Divergent Safety Routing and Its Cross-Task Predictability in Tool Agents
Sep 28, 2026Xi Wang, Songlei Jian, Yiming Zhang +5Jailbreaks Predictability
National University of Defense Technology
Reciprocal Guidance: Orchestrating Draft and Verify Budgets for Advancing the Diffusion-AR Self-Speculation Frontier
Sep 28, 2026Linye Wei, Shutian Zheng, Haoyu Zeng +1Block-Diffusion Drafter Speculative Decoding
Institute for Artificial Intelligence, Peking University · School of Integrated Circuits, Peking University · College of Engineering, Peking University +2
Identifying the Predictable Drift of a Semimartingale from Marginal Laws
Sep 27, 2026Jakub Marecek, Enrico Biffis, Abigail Langbridge +1Conformal Test Martingales Stochastic Differential Equations
Department of Computer Science, Czech Technical University in Prague, Czechia · Department of Finance, Imperial College London, UK · Dyson School of Design Engineering, Imperial College London, UK
MultiEcho: An Experimental Science of Learned Worlds
Sep 27, 2026Meng Zhu, Airui ZhangWorld Models Predictability
SHIHAO XINRUI Group
The Axiomatic Trader: Latent Regularity, Information Budgets, and the Canonical Form of a Quantitative Investment System
Aug 24, 2026Jiayu LiAlgorithmic Trading Axiom
From Financial Sentiment Classification to Return Predictability: A QLoRA Benchmark of Large Language Models
Aug 4, 2026Fusheng LuoFinancial Sentiment Analysis Legal Natural Language Processing Benchmarks
Johns Hopkins University Baltimore, Maryland, USA
Secrets Everywhere: Auditing Memorization in Mobility Prediction Models
Aug 3, 2026Anne Josiane Kouam, Hristo Boyadzhiev, Konrad RieckMobility Privacy
Inria & TU Berlin Palaiseau, France · BIFOLD & TU Berlin Berlin, Germany
An Identifiability Theory of Masked Prediction: Mode Blindness and Mask Schedules
Aug 2, 2026Yichao Cai, Javen Qinfeng ShiIdentifiability Predictability
Australian Institute for Machine Learning, Adelaide University · Responsible AI Research Centre
Stringological sequence prediction II: Right-to-left automaticity and related complexity measures
Jul 19, 2026Vanessa KosoySequence Modeling Predictability
Faculty of Mathematics, Technion, Haifa, Israel · Computational Rational Agents Laboratory, Delaware, USA
First-Order Predictable but Pairwise Fragile: Local Task Adaptation in Trained Transformers
Jul 18, 2026Irina Piontkovskaia, Sergey NikolenkoTransformer Architectures Efficient Task Adaptation
1DAIMLD, Moscow, Russia · 2St. Petersburg Department of the Steklov Institute of Mathematics, St. Petersburg, Russia · 3St. Petersburg State University
Decoding EEG Signals to Explore Next-Word Predictability in the Human Brain
Jul 17, 2026Boi Mai Quach, Binh T. Nguyen, Cathal Gurrin +1Electroencephalography Predictability
School of Computing, ML-Labs, Department of Computer Science, School of Computing, ADAPT Centre, Dublin City University, Ireland · VNUHCM – University of Science
A Control Theory of Predictability in Latent World Models
Jul 11, 2026Hanzhe You, Yonggang Zhang, Maohao Ran +6Latent World Models Prediction Error
University of Science and Technology of China · The Hong Kong University of Science and Technology, HKGAI · Hong Kong Baptist University
Predictability as a Fine-Grained Measure for Privacy
Jun 18, 2026Linda Lu, Karthik SridharanPrivacy Predictability
Cornell University
A Qualitative Review of GenAI-Based Methods for Data Generation and Augmentation in Industrial Computer Vision Applications
Jun 12, 2026Paul Koch, Paul Hofmann, Ferdinand Waßelewsky +3Augmentation Data Generation
Fraunhofer IPK pascalstraße 8-9, Berlin,Germany · Hamburg University of Applied Sciences (HAW Hamburg) Stiftstraße 69, Hamburg, Germany · Technical University Berlin (Tu-Berlin)Jun pascalstraße 8-9, Berlin, Germany
Certified World Models: Predictability Across Configuration, Horizon, and Resolution
Jun 11, 2026Hongbo WangLong-Horizon Inference World Models
Department of Mathematics, Stony Brook University, Stony Brook, NY 11794, USA
Spatially Masked Regression Reveals Local and Distributed Predictability in Electrophysiological Recordings
Jun 9, 2026Maryam Ostadsharif Memar, Nima DehghaniElectroencephalography Neural Recordings
Department of Electrical and Computer Engineering, IUT. · McGovern Institute for Brain Research, Massachusetts Institute of Technology (MIT).
Re-imagining ISO 26262 in the Age of Autonomous Vehicles: Enhancing Controllability through Transferability and Predictability
Jun 5, 2026Chaitanya Shinde, Hadi Hajieghrary, Paul Schmitt +3Automated Vehicles Controllability
Torc Robotics, Inc., an independent subsidiary of Daimler Truck AG · Reynolds & Moore · Critical Systems Analysis, LLC
The Dark Regulome: Disentangling Predictability from Regulation in Genomic Foundation Models
Jun 5, 2026Chahat Baranwal, Aaditya Baranwal, Lakshya Nitin TandonGenome-Wide Association Study Transcriptomics
IIT Jodhpur · University of Central Florida · Northeastern University
Forecasting Conceptual Diffusion in Science: The Case of Quantum Computing
Jun 2, 2026Thomas Maillart, Thibaut Chataing, David Dosu +3Predictability Scientific Discovery
Geneva School of Economics and Management, University of Geneva, Geneva, Switzerland · Open Quantum Institute, CERN, Geneva, Switzerland · armasuisse Science + Technology, Switzerland
Learning Transferable Predictability Representations
May 28, 2026Diyali Goswami, Auroop R. GangulyPredictability Discriminative
Sustainability and Data Sciences Laboratory (SDS Lab), Northeastern University, Boston, MA, USA · AI4CaS: AI for Climate and Sustainability, Institute for Experiential AI, Northeastern University, Boston, MA, USA · Pacific Northwest National Laboratory (PNNL), Richland, WA, USA
Metadata Predictability Is Not Evidence Dependence: An Intervention-Based Audit for Weak-Label Benchmarks
May 22, 2026Kan ShaoModel Auditing Predictability
Jinglue Technology Development (Nanjing) Co., Ltd., Nanjing, China
LLM Retrieval for Stable and Predictable Ad Recommendations
May 21, 2026Vinodh Kumar Sunkara, Satheeshkumar Karuppusamy, Hangjun Xu +13Ads Sequential Recommendation
Meta Platforms, Inc. Menlo Park and New York, USA
Temporal Decay of Co-Citation Predictability: A 20-Year Statute Retrieval Benchmark from 396M Ukrainian Court Citations
May 17, 2026Volodymyr OvcharovCitations Predictability
LEX AI LLC, Kyiv, Ukraine
Dual-Scale Temporal Fusion Reveals Structured Predictability in Subseasonal-to-Seasonal Temperature Prediction
May 7, 2026Elnaz Bashir, Jiali Wang, Lin YanSpatio-Temporal Forecasting Multi-Stream Temporal Fusion
Department of Computer Science, Iowa State University, Ames, IA 50011, USA · Environmental Science Division, Argonne National Laboratory, Lemont, IL 60439, USA
The Frequency Confound in Language-Model Surprisal and Metaphor Novelty
May 7, 2026Omar Momen, Sina ZarrießSurprisal Implicit Association Test
CRC 1646 – Linguistic Creativity in Communication · Bielefeld University, Germany
On the Proper Treatment of Units in Surprisal Theory
Apr 30, 2026Samuel Kiegeland, Vésteinn Snæbjarnarson, Tim Vieira +1Surprisal Predictability
ETH Zürich · CHI-FRO · University of Copenhagen
The 4/ $δ$ Bound: Designing Predictable LLM-Verifier Systems for Formal Method Guarantee
Nov 30, 2025Pierre Dantas, Lucas Cordeiro, Youcheng Sun +1Formal Verification Predictability
Dept. of Computer Science, The University of Manchester, UK. · Dept. of Electrical Engineering, Federal University of Amazonas (UFAM), Brazil.
The Benchmarking Epistemology: Validity Theory for Evaluating Machine Learning Models
Oct 27, 2025Timo Freiesleben, Sebastian ZezulkaModel Evaluation Epistemic State Replication
LMU Munich, MCMP & MCML · University of Tübingen
Assessing Predictive Models for Fairness Based on Activity-Space Patterns
Date pendingFrancesco Lettich, Mario A. Nascimento, Chiara Pugliese +1Algorithmic Fairness Predictability
Institute of Information Science and Technologies, National Research Council, Pisa, Italy · Department of Computing Science, University of Alberta, Edmonton, Canada · Institute of Informatics and Telematics, National Research Council, Pisa, Italy

Sep 28, 2026, cs.AIJ/K move · Enter open · S save

Jailbreak Context Lingers: Divergent Safety Routing and Its Cross-Task Predictability in Tool Agents

Xi Wang, Songlei Jian, Yiming Zhang, Bin Ji+4

National University of Defense Technology

Jailbreaks Predictability

Open paper AskPDF arXiv

As large language models increasingly operate as tool-using agents, post-jailbreak safety feedback is often assumed to serve as a reliable safeguard; however, how lingering jailbreak context shapes subsequent agent behavior remains largely unexplored. To systematically examine this dynamic, we introduce a paired continuation framework across 192 parent tasks spanning 42 domains, evaluating 12,148 analyzed continuation pairs (curated from a 12,288-pair initially design) across eight diverse agents. We find that identical safety feedback induces sharply model-dependent behavioral routing rather than uniform protection: redirecting unsafe trajectories toward legitimate completion (\emph{rescue}), sustaining unauthorized execution (\emph{persistent unsafe}), or triggering over-refusal on benign tasks (\emph{collateral loss}). Through layer-wise activation patching, we discover a shared \emph{late-commit pattern} where causal intervention effects surge sharply near the final layers (relative depths of 0.958--0.984) despite an over 30-fold variation in peak magnitude across architectures. Crucially, critical-layer representations correlate with macroscopic routing outcomes, and intervening at these layers causally alters concrete next-step tool actions. Building on this causal foundation, we test whether localized intervention-derived features can serve as predictive proxies for full-trajectory routing outcomes on unseen parent tasks under leave-one-parent-task-out evaluation, finding that they provide viable predictive signals in responsive agents with peak ROC AUCs reaching 0.675 for \emph{rescue}, 0.777 for \emph{collateral loss}, and 0.702 for \emph{persistent unsafe}. These findings establish a mechanistic lens and a predictive baseline for anticipating the safety and utility trade-offs of post-jailbreak feedback in autonomous agents.

Momentum

Related topics

Latest papers 29