2/13/2026
ULTRA:Urdu Language Transformer-based Recommendation Architecture
Urdu language transformer-based recommendation system for personalized news retrieval in a low-resource language context.
2/13/2026
Improving Neural Retrieval with Attribution-Guided Query Rewriting
Attribution-guided query rewriting method that uses token-level explanations to improve neural retrieval system robustness.
2/13/2026
Zooming without Zooming: Region-to-Image Distillation for Fine-Grained Multimodal Perception
Region-to-Image Distillation: technique to improve fine-grained visual perception in multimodal LLMs without repeated zooming operations.
2/13/2026
Towards Fair and Comprehensive Evaluation of Routers in Collaborative LLM Systems
RouterXBench: evaluation framework for routers in collaborative LLM systems that deploy smaller models locally and offload complex queries to cloud services.
2/13/2026
Agentic AI for Cybersecurity: A Meta-Cognitive Architecture for Governable Autonomy
Meta-cognitive architecture for agentic AI in cybersecurity enabling accountable autonomous decision-making under adversarial uncertainty.
2/13/2026
Mitigating Mismatch within Reference-based Preference Optimization
Method to improve Direct Preference Optimization for LLM alignment by addressing issues when reference model prefers rejected responses.
2/13/2026
Leveraging LLMs to support co-evolution between definitions and instances of textual DSLs: A Systematic Evaluation
Systematic evaluation of using LLMs to co-evolve domain-specific language definitions and instances during grammar changes.
2/13/2026
Who Does What? Archetypes of Roles Assigned to LLMs During Human-AI Decision-Making
Study of interaction patterns and role assignment between humans and LLMs in collaborative high-stakes decision-making systems.
2/13/2026
AdaptEvolve: Improving Efficiency of Evolutionary AI Agents through Adaptive Model Selection
Framework for dynamically selecting among LLMs in evolutionary agents to balance computational efficiency and reasoning capability.
2/13/2026
Manifold-Aware Temporal Domain Generalization for Large Language Models
Manifold-aware geometric approach for temporal domain generalization in LLMs under parameter-efficient constraints.
2/13/2026
Accelerating Robotic Reinforcement Learning with Agent Guidance
Agent guidance approach accelerates robotic reinforcement learning without human-in-the-loop bottlenecks.
2/13/2026
Evaluating AGENTS.md: Are Repository-Level Context Files Helpful for Coding Agents?
Evaluation of context files (AGENTS.md) effectiveness for improving coding agent task completion on real repositories.
2/13/2026
On the Sensitivity of Firing Rate-Based Federated Spiking Neural Networks to Differential Privacy
Analysis of differential privacy impact on federated spiking neural networks in distributed neuromorphic learning.
2/13/2026
An Empirical Study of the Imbalance Issue in Software Vulnerability Detection
Study of class imbalance effects on deep learning vulnerability detection in source code.
2/13/2026
Fourier Transformers for Latent Crystallographic Diffusion and Generative Modeling
Fourier-based generative pipeline for modeling crystalline materials using diffusion models with physical constraints.
2/13/2026
ModelWisdom: An Integrated Toolkit for TLA+ Model Visualization, Digest and Repair
ModelWisdom toolkit for TLA+ model visualization and automated repair of formal specification counterexamples.
2/13/2026
Choose Your Agent: Tradeoffs in Adopting AI Advisors, Coaches, and Delegates in Multi-Party Negotiation
Behavioral study examining user preferences for different LLM assistant modalities in multi-party negotiation scenarios.
2/13/2026
DeepSight: An All-in-One LM Safety Toolkit
DeepSight comprehensive toolkit for LLM and MLLM safety including evaluation, diagnosis, and alignment workflows.
2/13/2026
Multi Graph Search for High-Dimensional Robot Motion Planning
Multi Graph Search algorithm for efficient motion planning in high-dimensional robotic systems.
2/13/2026
On the Complexity of Offline Reinforcement Learning with $Q^\star$-Approximation and Partial Coverage
Theoretical analysis of offline reinforcement learning under Q-approximation and partial coverage conditions.
2/13/2026
Meta-Sel: Efficient Demonstration Selection for In-Context Learning via Supervised Meta-Learning
Meta-Sel uses supervised meta-learning for efficient demonstration selection in in-context learning with computational budget constraints.
2/13/2026
Learning beyond Teacher: Generalized On-Policy Distillation with Reward Extrapolation
On-policy distillation with reward extrapolation extends KL-constrained RL to improve student model performance beyond teacher alignment.
2/13/2026
On the Adoption of AI Coding Agents in Open-source Android and iOS Development
Empirical study of AI coding agent adoption analyzing 2,901 PR contributions to open-source Android and iOS projects.
2/13/2026
dVoting: Fast Voting for dLLMs
dVoting introduces fast voting mechanism for Diffusion LLMs enabling parallel token generation and efficient test-time scaling.
2/13/2026
3DGSNav: Enhancing Vision-Language Model Reasoning for Object Navigation via Active 3D Gaussian Splatting
3DGSNav enhances vision-language models for zero-shot object navigation in embodied AI agents using 3D Gaussian splatting.
2/13/2026
SAGEO Arena: A Realistic Environment for Evaluating Search-Augmented Generative Engine Optimization
SAGEO Arena environment evaluates Search-Augmented Generative Engines and optimization techniques for web content visibility in AI responses.
2/13/2026
Visual Reasoning Benchmark: Evaluating Multimodal LLMs on Classroom-Authentic Visual Problems from Primary Education
VRB benchmark dataset for evaluating multimodal LLMs on visual reasoning tasks from primary education math problems.
2/13/2026
DeepGen 1.0: A Lightweight Unified Multimodal Model for Advancing Image Generation and Editing
DeepGen 1.0: lightweight 5B parameter unified multimodal model for competitive image generation and editing capabilities.
2/13/2026
The Observer Effect in World Models: Invasive Adaptation Corrupts Latent Physics
Studies whether neural world models internalize true physics or exploit shortcuts, proposing methods to avoid measurement corruption.
2/13/2026
Towards On-Policy SFT: Distribution Discriminant Theory and its Applications in LLM Training
Distribution Discriminant Theory enables on-policy supervised fine-tuning for LLMs, bridging SFT-RL generalization gap.
2/13/2026
Energy-Aware Spike Budgeting for Continual Learning in Spiking Neural Networks for Neuromorphic Vision
Energy-aware continual learning method for spiking neural networks on neuromorphic vision systems with event-based cameras.
2/13/2026
Olmix: A Framework for Data Mixing Throughout LM Development
Olmix: framework for determining optimal data mixing ratios during language model training with practical design guidance.
2/13/2026
ExtractBench: A Benchmark and Evaluation Methodology for Complex Structured Extraction
ExtractBench: benchmark and methodology for evaluating LLM-based PDF-to-JSON structured data extraction at enterprise scale.
2/13/2026
A technical curriculum on language-oriented artificial intelligence in translation and specialised communication
Technical curriculum on language-oriented AI literacy for translation and specialized communication industry professionals.
2/13/2026
On the implicit regularization of Langevin dynamics with projected noise
Studies implicit regularization in Langevin dynamics with projected noise for understanding stochastic gradient descent in symmetric models.
2/13/2026
AttentionRetriever: Attention Layers are Secretly Long Document Retrievers
AttentionRetriever leverages LLM attention layers for long document retrieval, enabling better RAG without separate retrieval models.
2/13/2026
UniT: Unified Multimodal Chain-of-Thought Test-time Scaling
UniT enables unified multimodal models to iteratively refine outputs through test-time scaling with chain-of-thought reasoning.
2/13/2026
Scaling Verification Can Be More Effective than Scaling Policy Learning for Vision-Language-Action Alignment
Vision-Language-Action model test-time verification reduces instruction-action misalignment more effectively than policy scaling.
2/13/2026
Can Complexity and Uncomputability Explain Intelligence? SuperARC: A Test for Artificial Super Intelligence Based on Recursive Compression
SuperARC: AGI/ASI test based on recursive compression and mathematical randomness rather than human-centric questions.
2/13/2026
SKATE, a Scalable Tournament Eval: Weaker LLMs differentiate between stronger ones using verifiable challenges
SKATE: LLM evaluation framework where models generate and solve verifiable tasks for each other, treating evaluation as a competitive game.
2/13/2026
Leveraging Generative AI for Human Understanding: Meta-Requirements and Design Principles for Explanatory AI as a new Paradigm
Proposes design principles for explainable AI systems that integrate domain knowledge and contextual reasoning for better human comprehension.
2/13/2026
Hybrid Reinforcement Learning and Search for Flight Trajectory Planning
Combines reinforcement learning with search-based path planners to optimize flight trajectory planning for emergency route recalculation.
2/13/2026
Roundtable Policy: Confidence-Weighted-Consensus Aggregation Improves Multi-Agent-System Reasoning
Multi-agent system reasoning improvement through confidence-weighted consensus aggregation, inspired by democratic committee decision-making.
2/13/2026
Logical Structure as Knowledge: Enhancing LLM Reasoning via Structured Logical Knowledge Density Estimation
Method for improving LLM reasoning by optimizing training data logical complexity and entailment relationships rather than just factuality or diversity.
2/13/2026
Human Behavior Atlas: Benchmarking Unified Psychological and Social Behavior Understanding
Human Behavior Atlas: Unified benchmark dataset for understanding psychological and social behaviors through multimodal AI, covering affective and cognitive states.
2/13/2026
MARSHAL: Incentivizing Multi-Agent Reasoning via Self-Play with Strategic LLMs
MARSHAL: Reinforcement learning framework for training LLMs in multi-agent competition and cooperation using self-play and strategic reasoning.
2/13/2026
SMaRT: Select, Mix, and ReinvenT -- A Strategy Fusion Framework for LLM-Driven Reasoning and Planning
SMaRT: Strategy fusion framework combining multiple reasoning approaches for LLM task automation, maximizing performance and robustness over single-strategy methods.
2/13/2026
Evaluating LLM Reasoning Beyond Correctness and CoT
SIEV framework: Evaluation method for LLM reasoning beyond correctness, modeling reasoning as dynamic dialectical trajectory rather than static chains.
2/13/2026
DriveSafe: A Hierarchical Risk Taxonomy for Safety-Critical LLM-Based Driving Assistants
DriveSafe: Domain-specific risk taxonomy and evaluation framework for LLM-based driving assistants addressing safety-critical automotive scenarios.
2/13/2026
Phase Transition for Budgeted Multi-Agent Synergy
Theoretical analysis of multi-agent system performance under fixed inference budget; predicts regimes where agents help, saturate, or fail based on context windows and communication constraints.