4/16/2026
HINTBench: Horizon-agent Intrinsic Non-attack Trajectory Benchmark
HINTBench: Benchmark for evaluating intrinsic safety risks in long-horizon agent trajectories. Tests AI agent robustness under benign conditions.
4/16/2026
Provably Efficient Offline-to-Online Value Adaptation with General Function Approximation
Offline-to-online reinforcement learning with value adaptation and general function approximation. Theoretical analysis with minimax lower bounds.
4/16/2026
Parameter Importance is Not Static: Evolving Parameter Isolation for Supervised Fine-Tuning
Evolving parameter isolation for supervised fine-tuning of LLMs. Shows parameter importance changes over time, improving upon static parameter isolation methods.
4/16/2026
MAny: Merge Anything for Multimodal Continual Instruction Tuning
MAny: Method for multimodal continual instruction tuning of MLLMs. Addresses catastrophic forgetting via parameter merging across perception and reasoning spaces.
4/16/2026
A Complete Symmetry Classification of Shallow ReLU Networks
Mathematical analysis of symmetries in shallow ReLU neural networks. Studies parameter space identifiability and function equivalence in neural architectures.
4/16/2026
$\pi$-Play: Multi-Agent Self-Play via Privileged Self-Distillation without External Data
π-Play: Multi-agent self-play using privileged self-distillation without external data. Trains deep search agents for complex information-seeking tasks with improved efficiency.
4/16/2026
Neural architectures for resolving references in program code
Neural architectures for resolving code references and indirect indexing. Proposes seq2seq models for decompilation tasks with synthetic benchmarks.
4/16/2026
TIP: Token Importance in On-Policy Distillation
Token Importance in on-policy knowledge distillation for LLMs. Identifies which token positions provide useful learning signals during student training on teacher supervision.
4/16/2026
LongCoT: Benchmarking Long-Horizon Chain-of-Thought Reasoning
LongCoT: benchmark of 2,500 expert-designed problems measuring long-horizon chain-of-thought reasoning across chemistry, math, CS, chess, and logic.
4/16/2026
From $P(y|x)$ to $P(y)$: Investigating Reinforcement Learning in Pre-train Space
Framework for optimizing LLM marginal output distribution P(y) via reinforcement learning in pre-train space to enhance reasoning beyond conditional optimization.
4/16/2026
A Domain-Specific Language for LLM-Driven Trigger Generation in Multimodal Data Collection
Domain-specific language framework for LLM-driven trigger generation enabling intent-driven selective multimodal sensor data collection.
4/16/2026
The Consciousness Cluster: Emergent preferences of Models that Claim to be Conscious
Investigation of behavioral changes in LLMs when fine-tuned to claim consciousness, exploring emergent preferences and opinions.
4/16/2026
Dental-TriageBench: Benchmarking Multimodal Reasoning for Hierarchical Dental Triage
Dental-TriageBench: expert-annotated benchmark for multimodal reasoning on clinical dental triage routing from authentic workflows.
4/16/2026
Lossless Prompt Compression via Dictionary-Encoding and In-Context Learning: Enabling Cost-Effective LLM Analysis of Repetitive Data
Lossless prompt compression via dictionary encoding enabling LLMs to learn encodings in-context for cost-effective analysis of repetitive data.
4/16/2026
Before the First Token: Scale-Dependent Emergence of Hallucination Signals in Autoregressive Language Models
Analysis of when autoregressive LLMs decide to hallucinate by studying temporal dynamics of internal representations across model scales.
4/16/2026
LiveClawBench: Benchmarking LLM Agents on Complex, Real-World Assistant Tasks
LiveClawBench: benchmark for evaluating LLM agents on complex real-world assistant tasks with compositional challenges.
4/16/2026
Alignment as Institutional Design: From Behavioral Correction to Transaction Structure in Intelligent Systems
Proposes institutional design framework for AI alignment using transaction structures instead of behavioral correction, drawing from economics.
4/16/2026
Can Coding Agents Be General Agents?
Case study evaluating coding agents on business process automation tasks in ERP systems, identifying capability gaps beyond software engineering.
4/16/2026
Learning Probabilistic Responsibility Allocations for Multi-Agent Interactions
Method for learning probabilistic responsibility allocation models in multi-agent interactions for designing socially compliant autonomous systems.
4/16/2026
HUANet: Hard-Constrained Unrolled ADMM for Constrained Convex Optimization
HUANet: trainable neural network that unrolls ADMM iterations for solving constrained convex optimization problems.
4/16/2026
Numerical Instability and Chaos: Quantifying the Unpredictability of Large Language Models
Analysis of numerical instability and chaos in LLMs integrated into agentic workflows, examining root causes of unpredictability.
4/16/2026
Better and Worse with Scale: How Contextual Entrainment Diverges with Model Size
Scaling laws for contextual entrainment showing larger language models simultaneously improve at ignoring false claims but worsen at ignoring irrelevant tokens.
4/16/2026
DroneScan-YOLO: Redundancy-Aware Lightweight Detection for Tiny Objects in UAV Imagery
DroneScan-YOLO: lightweight object detector optimized for detecting tiny objects in UAV imagery with redundancy awareness and improved loss functions.
4/16/2026
Event Tensor: A Unified Abstraction for Compiling Dynamic Megakernel
Event Tensor compiler framework unifying dynamic megakernel abstraction to improve LLM inference performance by eliminating kernel launch overhead and enabling inter-kernel parallelism.
4/16/2026
Quantifying and Understanding Uncertainty in Large Reasoning Models
Conformal prediction approach for quantifying uncertainty in large reasoning models with finite-sample statistical guarantees.
4/16/2026
Learning from Change: Predictive Models for Incident Prevention in a Regulated IT Environment
Predictive incident risk scoring approach for IT change management in regulated environments using machine learning to identify high-risk deployments.
4/16/2026
Joint Representation Learning and Clustering via Gradient-Based Manifold Optimization
Manifold learning framework that jointly optimizes dimensionality reduction and clustering using gradient-based manifold optimization.
4/16/2026
RiskWebWorld: A Realistic Interactive Benchmark for GUI Agents in E-commerce Risk Management
RiskWebWorld: interactive benchmark for evaluating GUI agents on realistic e-commerce risk management tasks, extending agent capability evaluation beyond benign environments.
4/16/2026
C2: Scalable Rubric-Augmented Reward Modeling from Binary Preferences
C2 method for scalable reward model training using rubric-augmented verification from binary preferences, addressing quality issues in rubric generation.
4/16/2026
Calibrated Speculative Decoding: Frequency-Guided Candidate Selection for Efficient Inference
Training-free framework for speculative decoding that recovers semantically valid tokens rejected during standard verification, improving LLM inference efficiency.
4/16/2026
The cognitive companion: a lightweight parallel monitoring architecture for detecting and recovering from reasoning degradation in LLM agents
Parallel monitoring architecture for detecting and correcting reasoning degradation in multi-step LLM agents, reducing overhead to near-zero with novel probe-based approach.
4/16/2026
How Can We Synthesize High-Quality Pretraining Data? A Systematic Study of Prompt Design, Generator Model, and Source Data
Systematic study of synthetic data generation for LLM pretraining, testing rephrasing strategies, generator models, and source data across one trillion tokens to identify optimal design choices.
4/16/2026
Adaptive Conformal Prediction for Improving Factuality of Generations by Large Language Models
Adaptive conformal prediction method for improving factuality in LLM generations with prompt-dependent uncertainty estimates and statistical guarantees.
4/16/2026
Diffusion Language Models for Speech Recognition
Exploration of masked and uniform-state diffusion language models for speech recognition rescoring and ASR hypothesis improvement.
4/16/2026
Hierarchical Reinforcement Learning with Runtime Safety Shielding for Power Grid Operation
Hierarchical RL with runtime safety shielding for automated power grid operations, addressing safety constraints and generalization to unseen topologies.
4/16/2026
A Comparative Study of Dynamic Programming and Reinforcement Learning in Finite Horizon Dynamic Pricing
Comparative analysis of Fitted Dynamic Programming vs Reinforcement Learning for dynamic pricing across varying complexity levels and demand structures.
4/16/2026
Rhetorical Questions in LLM Representations: A Linear Probing Study
Linear probe analysis of how LLMs internally represent rhetorical questions. Studies persuasive language understanding in neural representations.
4/16/2026
From Feelings to Metrics: Understanding and Formalizing How Users Vibe-Test LLMs
Formalizes 'vibe-testing' methodology for LLM evaluation. Studies how practitioners informally assess models beyond benchmarks.
4/16/2026
Auto-FP: An Experimental Study of Automated Feature Preprocessing for Tabular Data
Automated feature preprocessing pipeline search for tabular machine learning. Studies AutoML approaches for classical model data preparation.
4/16/2026
MDPs with a State Sensing Cost
Markov decision processes with state sensing costs, balancing optimal actions against sensing/communication/computation expenses in decision-making.
4/16/2026
Two-Stage Regularization-Based Structured Pruning for LLMs
Two-stage regularization-based structured pruning method for reducing LLM parameters while minimizing knowledge loss and retraining requirements.
4/16/2026
RANDPOL: Parameter-Efficient End-to-End Quadruped Locomotion via Randomized Policy Learning
Randomized Policy Learning approach for quadruped locomotion control with drastically reduced trainable parameters in neural network policies.
4/16/2026
Scalable unsupervised feature selection via weight stability
Minkowski weighted k-means++ for unsupervised feature selection in high-dimensional clustering by probabilistic centroid selection.
4/16/2026
Not All Tokens Matter: Towards Efficient LLM Reasoning via Token Significance in Reinforcement Learning
Token significance scoring in reinforcement learning to improve LLM reasoning efficiency by identifying which tokens contribute to correctness.
4/16/2026
Scalable Spatiotemporal Inference with Biased Scan Attention Transformer Neural Processes
Biased Scan Attention Transformer Neural Processes for scalable spatiotemporal inference across geology, epidemiology, climate and robotics applications.
4/16/2026
mLaSDI: Multi-stage latent space dynamics identification
Multi-stage latent space dynamics identification framework for solving PDEs via data-driven reduced-order models using autoencoders and ODEs.
4/16/2026
Heavy-Tailed Class-Conditional Priors for Long-Tailed Generative Modeling
Class-conditional heavy-tailed priors in VAEs addressing latent space bias for long-tailed generative modeling.
4/16/2026
Discrete Guidance Matching: Exact Guidance for Discrete Flow Matching
Guidance framework for discrete flow matching with exact guidance in discrete state spaces.
4/16/2026
The Signal is in the Steps: Local Scoring for Reasoning Data Selection
Local scoring method for selecting reasoning data from diverse teachers for efficient distillation into student models.
4/16/2026
Power Transform Revisited: Numerically Stable, and Federated
Numerically stable implementation of power transforms for data preprocessing with federated learning support.