2/16/2026
CoPE-VideoLM: Codec Primitives For Efficient Video Language Models
CoPE-VideoLM uses codec primitives for efficient video understanding in language models, reducing computational overhead while preserving temporal details.
2/16/2026
Semantic Chunking and the Entropy of Natural Language
Statistical model capturing multi-scale structure of natural language entropy, benchmarking LLM compression rates against information-theoretic limits.
2/16/2026
SaVe-TAG: LLM-based Interpolation for Long-Tailed Text-Attributed Graphs
SaVe-TAG uses LLMs to interpolate long-tailed text-attributed graphs, improving GNN generalization across head and tail classes through semantic preservation.
2/16/2026
Mathematics and Machine Creativity: A Survey on Bridging Mathematics with AI
Survey of AI applications in mathematics including reinforcement learning and LLMs advancing mathematical research and discovery.
2/16/2026
SCAN: Semantic Document Layout Analysis for Textual and Visual Retrieval-Augmented Generation
SCAN analyzes semantic document layout for retrieval-augmented generation using VLMs, improving RAG performance on information-dense documents.
2/16/2026
AutoGPS: Automated Geometry Problem Solving via Multimodal Formalization and Deductive Reasoning
AutoGPS neuro-symbolic framework solves geometry problems combining multimodal comprehension with deductive reasoning for improved reliability and interpretability.
2/16/2026
How to Train Your LLM Web Agent: A Statistical Diagnosis
Statistical diagnosis of LLM web agents identifies bottlenecks in multi-step interactions and proposes methods reducing compute costs for open-source systems.
2/16/2026
A Survey on Hypergame Theory: Modeling Misaligned Perceptions and Nested Beliefs for Multi-agent Systems
Survey on hypergame theory for multi-agent systems modeling misaligned perceptions and nested beliefs under uncertainty and incomplete information.
2/16/2026
EvoCut: Strengthening Integer Programs via Evolution-Guided Language Models
EvoCut automates generation of acceleration cuts for integer programming solvers using evolution-guided language models at symbolic modeling level.
2/16/2026
Difficulty-Aware Agentic Orchestration for Query-Specific Multi-Agent Workflows
DAAO dynamically orchestrates multi-agent LLM workflows by routing queries to appropriate agents based on difficulty, balancing efficiency and performance.
2/16/2026
VoiceAgentBench: Are Voice Assistants ready for agentic tasks?
VoiceAgentBench benchmark evaluates speech language models on agentic tasks and adversarial robustness beyond isolated capabilities like transcription.
2/16/2026
RLIE: Rule Generation with Logistic Regression, Iterative Refinement, and Evaluation for Large Language Models
RLIE framework integrates LLMs with probabilistic rule learning to generate weighted natural language rules, coupling rule interactions with probabilistic inference.
2/16/2026
Agentic AI Security: Threats, Defenses, Evaluation, and Open Challenges
Comprehensive survey of security threats, defenses, and evaluation methods specific to autonomous agentic AI systems with tool use.
2/16/2026
Impact of Data-Oriented and Object-Oriented Design on Performance and Cache Utilization with Artificial Intelligence Algorithms in Multi-Threaded CPUs
Performance analysis comparing data-oriented vs object-oriented design paradigms for AI algorithms on multi-threaded CPUs.
2/16/2026
TA-KAND: Two-stage Attention Triple Enhancement and U-KAN based Diffusion For Few-shot Knowledge Graph Completion
Knowledge graph completion technique using attention enhancement and diffusion models for few-shot long-tail relations.
2/16/2026
Understanding Chain-of-Thought in Large Language Models via Topological Data Analysis
Topological data analysis framework to understand why different chain-of-thought reasoning paths perform differently in LLMs.
2/16/2026
Finetuning Large Language Models for Automated Depression Screening in Nigerian Pidgin English: GENSCORE Pilot Study
Fine-tuned LLM for depression screening in Nigerian Pidgin English addresses cultural and linguistic accessibility gaps.
2/16/2026
Panning for Gold: Expanding Domain-Specific Knowledge Graphs with General Knowledge
Methods for enriching domain-specific knowledge graphs by integrating general knowledge graphs systematically.
2/16/2026
Preventing the Collapse of Peer Review Requires Verification-First AI
Proposes verification-first approach to AI-assisted peer review using truth-coupling metric instead of review-mimicking.
2/16/2026
ATLAS : Adaptive Self-Evolutionary Research Agent with Task-Distributed Multi-LLM Supporters
Self-evolving research agent framework with task-distributed multi-LLM supporters for adaptive long-horizon problem-solving.
2/16/2026
WideSeek-R1: Exploring Width Scaling for Broad Information Seeking via Multi-Agent Reinforcement Learning
Multi-agent reinforcement learning system exploring width scaling for broad information seeking across multiple LLM agents.
2/16/2026
FiMI: A Domain-Specific Language Model for Indian Finance Ecosystem
Domain-specialized financial language model for Indian digital payment systems adapted from Mistral architecture.
2/16/2026
PuYun-LDM: A Latent Diffusion Model for High-Resolution Ensemble Weather Forecasts
Latent diffusion model for high-resolution ensemble weather forecasting addressing diffusability challenges in meteorological data.
2/16/2026
When Should LLMs Be Less Specific? Selective Abstraction for Reliable Long-Form Text Generation
Selective abstraction framework for LLMs to reduce factual errors by strategically abstracting when confidence is low in long-form generation.
2/16/2026
CATP: Cross-Attention Token Pruning for Accuracy Preserved Multimodal Model Inference
Token pruning technique for multimodal models using cross-attention layers to improve inference efficiency and accuracy.
2/16/2026
Compressible Dynamics in Deep Overparameterized Low-Rank Learning & Adaptation
Methods to leverage low-dimensional structures in overparameterized models to reduce computational costs while maintaining benefits.
2/16/2026
LTSM-Bundle: A Toolbox and Benchmark on Large Language Models for Time Series Forecasting
Toolbox and benchmark for large time series models applying LLMs to time series forecasting with transformer-based approaches.
2/16/2026
Explaining and Mitigating the Modality Gap in Contrastive Multimodal Learning
Analysis of modality gap in contrastive multimodal learning models like CLIP and methods to improve cross-modal alignment.
2/16/2026
Hierarchical Retrieval at Scale: Bridging Transparency and Efficiency
Hierarchical retrieval methods for information retrieval in intelligent systems offering interpretability and efficiency trade-offs.
2/16/2026
Redefining Evaluation Standards: A Unified Framework for Evaluating the Korean Capabilities of Language Models
Unified evaluation framework for Korean LLM capabilities addressing reproducibility gaps and inconsistent benchmarking protocols.
2/16/2026
PlanetServe: A Decentralized, Scalable, and Privacy-Preserving Overlay for Democratizing Large Language Model Serving
Decentralized peer-to-peer overlay network for scalable LLM serving enabling small organizations to deploy models efficiently.
2/16/2026
Holistic Continual Learning under Concept Drift with Adaptive Memory Realignment
Continual learning framework addressing concept drift with adaptive memory realignment for dynamic data streams.
2/16/2026
Investigating Redundancy in Multimodal Large Language Models with Multiple Vision Encoders
Investigation of redundancy in multimodal LLMs with multiple vision encoders through systematic masking analysis.
2/16/2026
FISHER: A Foundation Model for Multi-Modal Industrial Signal Comprehensive Representation
Foundation model for multi-modal industrial signal analysis addressing heterogeneous SCADA system data with scaling laws.
2/16/2026
R-Zero: Self-Evolving Reasoning LLM from Zero Data
Self-evolving LLM that autonomously generates and learns from reasoning tasks without human-curated data or labels.
2/16/2026
MLLM-CTBench: A Benchmark for Continual Instruction Tuning with Reasoning Process Diagnosis
Benchmark for continual instruction tuning of multimodal LLMs across seven tasks with reasoning process diagnosis.
2/16/2026
ToolACE-MT: Non-Autoregressive Generation for Agentic Multi-Turn Interaction
Non-autoregressive generation method for multi-turn agentic interactions with LLMs, enabling efficient synthetic data generation for tool use.
2/16/2026
Diffusion-Based Scenario Tree Generation for Multivariate Time Series Prediction and Multistage Stochastic Optimization
Diffusion-based framework for generating scenario trees for multivariate time series prediction and stochastic optimization in energy/finance.
2/16/2026
Eliminating stability hallucinations in llm-based tts models via attention guidance
Method for eliminating stability hallucinations in LLM-based text-to-speech models using attention guidance and alignment scoring.
2/16/2026
Provable Training Data Identification for Large Language Models
Formal framework for training data identification in LLMs as set-level inference problem with statistical guarantees for copyright and privacy auditing.
2/16/2026
Low-Resource Dialect Adaptation of Large Language Models: A French Dialect Case-Study
Continual pre-training approach for adapting LLMs to low-resource French dialects using low-rank adaptation under tight compute budgets.
2/16/2026
MASPRM: Multi-Agent System Process Reward Model
Multi-Agent System Process Reward Model that guides multi-agent inference through value assignment to partial transcripts using Monte Carlo Tree Search.
2/16/2026
Watermarking Discrete Diffusion Language Models
Watermarking technique for discrete diffusion language models to track AI-generated content and differentiate from human creations.
2/16/2026
Reasoning about Intent for Ambiguous Requests
Research on handling ambiguous requests in LLMs by generating multiple interpretation-answer pairs trained with RL and custom reward functions.
2/16/2026
SGM: Safety Glasses for Multimodal Large Language Models via Neuron-Level Detoxification
SGM neuron-level detoxification method for multimodal LLMs, removing toxic signals at white-box intervention level to address safety risks from pretraining corpora.
2/16/2026
From Prompt to Product: A Human-Centered Benchmark of Agentic App Generation Systems
Human-centered benchmark evaluating prompt-to-app agentic AI systems generating full-stack web applications, measuring visual polish and functional correctness.
2/16/2026
Enhancing guidance for missing data in diffusion-based sequential recommendation
Diffusion-based sequential recommendation method addressing missing data through guidance mechanism that captures critical turning points in user interests.
2/16/2026
Privacy in Human-AI Romantic Relationships: Concerns, Boundaries, and Agency
Interview study examining privacy concerns and boundaries in human-AI romantic relationships facilitated by LLM-based applications across relationship stages.
2/16/2026
SAGE: Sequence-level Adaptive Gradient Evolution for Generative Recommendation
SAGE optimizer improves reinforcement learning-based preference optimization for generative recommendation systems, addressing symmetric conservatism in policy bounds.
2/16/2026
Beyond Static Question Banks: Dynamic Knowledge Expansion via LLM-Automated Graph Construction and Adaptive Generation
LLM-automated system for constructing knowledge graphs and generating adaptive educational questions, addressing scalability limitations of manual curation.