4/8/2026
JTON: A Token-Efficient JSON Superset with Zen Grid Tabular Encoding for Large Language Models
JTON format, JSON superset with Zen Grid encoding for token-efficient structured data processing in LLMs.
4/8/2026
Joint Knowledge Base Completion and Question Answering by Combining Large Language Models and Small Language Models
Joint knowledge base completion and QA using combined large and small language models for KB-related tasks.
4/8/2026
HybridKV: Hybrid KV Cache Compression for Efficient Multimodal Large Language Model Inference
KV cache compression technique for multimodal LLM inference, reducing memory overhead and latency with hybrid compression strategy.
4/8/2026
Context-Value-Action Architecture for Value-Driven Large Language Model Agents
Architecture for value-driven LLM agents addressing behavioral rigidity through context-value-action design.
4/8/2026
MARL-GPT: Foundation Model for Multi-Agent Reinforcement Learning
Foundation model enabling single GPT-based agent to perform across diverse multi-agent reinforcement learning tasks and environments.
4/8/2026
Towards Trustworthy Report Generation: A Deep Research Agent with Progressive Confidence Estimation and Calibration
Research agent framework for generating trustworthy reports with confidence estimation and calibration mechanisms.
4/8/2026
Beyond Compromise: Pareto-Lenient Consensus for Efficient Multi-Preference LLM Alignment
Multi-objective preference alignment for LLMs using Pareto-lenient consensus to handle diverse human values in model training.
4/8/2026
Flowr -- Scaling Up Retail Supply Chain Operations Through Agentic AI in Large Scale Supermarket Chains
AI agents for retail supply chain operations, automating demand forecasting, procurement, and inventory replenishment in supermarket chains.
4/8/2026
Epistemic Blinding: An Inference-Time Protocol for Auditing Prior Contamination in LLM-Assisted Analysis
Proposes epistemic blinding, an inference-time auditing protocol to separate memorized priors from data-driven inference in LLM-assisted agentic analysis systems.
4/8/2026
How LLMs Follow Instructions: Skillful Coordination, Not a Universal Mechanism
Investigates instruction-following mechanisms in LLMs through diagnostic probing, finding evidence for compositional skill deployment over universal mechanism.
4/8/2026
ACE-Bench: Agent Configurable Evaluation with Scalable Horizons and Controllable Difficulty under Lightweight Environments
Proposes ACE-Bench, agent evaluation benchmark with unified grid-based planning tasks, lightweight environments, and configurable difficulty/horizon control.
4/8/2026
Claw-Eval: Toward Trustworthy Evaluation of Autonomous Agents
Introduces Claw-Eval, an end-to-end evaluation suite for autonomous agents addressing trajectory-opaque grading, safety, and interaction modality coverage.
4/8/2026
Web Retrieval-Aware Chunking (W-RAC) for Efficient and Cost-Effective Retrieval-Augmented Generation Systems
Proposes Web Retrieval-Aware Chunking (W-RAC) for efficient RAG document chunking to balance retrieval quality, latency, and cost on web-scale content.
4/8/2026
TDA-RC: Task-Driven Alignment for Knowledge-Based Reasoning Chains in Large Language Models
Proposes Task-Driven Alignment (TDA-RC) for improving reasoning chains in LLMs by bridging logical gaps between CoT and multi-round thought paradigms.
4/8/2026
The Illusion of Latent Generalization: Bi-directionality and the Reversal Curse
Evaluates bidirectional training objectives (MLM, masked attention) to mitigate the reversal curse in autoregressive language models.
4/8/2026
Inclusion-of-Thoughts: Mitigating Preference Instability via Purifying the Decision Space
Introduces Inclusion-of-Thoughts (IoT), a strategy to reduce LLM instability on multiple-choice questions by filtering irrelevant distractors.
4/8/2026
SUMMIR: A Hallucination-Aware Framework for Ranking Sports Insights from LLMs
Proposes SUMMIR framework for ranking sports insights extracted by LLMs, addressing hallucinations with 7,900-article dataset across four sports.
4/8/2026
From PDF to RAG-Ready: Evaluating Document Conversion Frameworks for Domain-Specific Question Answering
Evaluates four open-source PDF-to-Markdown conversion frameworks (Docling, MinerU, Marker, DeepSeek OCR) for RAG document preprocessing impact on QA accuracy.
4/8/2026
Learning to Retrieve from Agent Trajectories
Studies how to design information retrieval systems for LLM agents versus humans, proposing learning-to-rank methods for agent trajectories.
4/8/2026
Generative AI for Video Trailer Synthesis: From Extractive Heuristics to Autoregressive Creativity
Surveys transition from heuristic-based to generative synthesis methods for automatic video trailer generation using LLMs and diffusion models.
4/8/2026
The Planetary Cost of AI Acceleration, Part II: The 10th Planetary Boundary and the 6.5-Year Countdown
Opinion piece on environmental and computational costs of scaling LLM agents and implications for planetary boundaries.
4/8/2026
MG$^2$-RAG: Multi-Granularity Graph for Multimodal Retrieval-Augmented Generation
Proposes MG²-RAG, a multi-granularity graph approach for retrieval-augmented generation in multimodal LLMs to improve cross-modal reasoning without costly text translation.
4/8/2026
Measuring the Permission Gate: A Stress-Test Evaluation of Claude Code's Auto Mode
Independent evaluation of Claude Code's auto mode permission system for AI coding agents, testing security gates on ambiguous authorization scenarios.
4/8/2026
Squeez: Task-Conditioned Tool-Output Pruning for Coding Agents
Introduces Squeez, a method for pruning tool outputs in coding agents by identifying minimal relevant evidence blocks. Includes 11,477-example benchmark from SWE-bench.
4/8/2026
CURE:Circuit-Aware Unlearning for LLM-based Recommendation
CURE enables privacy-preserving unlearning in LLM-based recommendation systems using circuit-aware techniques for removing user data.
4/8/2026
Cactus: Accelerating Auto-Regressive Decoding with Constrained Acceptance Speculative Sampling
Cactus improves speculative sampling for LLM decoding by relaxing strict distribution matching to allow acceptable variations like top-k sampling.
4/8/2026
Prune-Quantize-Distill: An Ordered Pipeline for Efficient Neural Network Compression
Prune-Quantize-Distill pipeline for neural network compression optimizing wall-clock inference time rather than parameter count or FLOPs.
4/8/2026
Architecture Without Architects: How AI Coding Agents Shape Software Architecture
Analysis of implicit architectural decisions made by AI coding agents, identifying five mechanisms and six prompt-architecture coupling patterns.
4/8/2026
FreakOut-LLM: The Effect of Emotional Stimuli on Safety Alignment
FreakOut-LLM framework investigates whether emotionally charged prompts compromise safety alignment in ten LLMs using psychological stimuli.
4/8/2026
Evaluation of Embedding-Based and Generative Methods for LLM-Driven Document Classification: Opportunities and Challenges
Comparative evaluation of embedding-based and generative models for document classification, showing Vision-Language Models with CoT achieve 82% zero-shot accuracy.
4/8/2026
PRIME: Prototype-Driven Multimodal Pretraining for Cancer Prognosis with Missing Modalities
PRIME enables multimodal self-supervised pretraining for cancer prognosis with missing modalities by combining histopathology, gene expression, and reports.
4/8/2026
Closed-Loop Autonomous Software Development via Jira-Integrated Backlog Orchestration: A Case Study in Deterministic Control and Safety-Constrained Automation
Case study of closed-loop software development system managing backlog via deterministic pipeline with Jira integration and safety constraints.
4/8/2026
EduIllustrate: Towards Scalable Automated Generation Of Multimodal Educational Content
EduIllustrate benchmark evaluates LLMs on generating multimodal educational content combining accurate diagrams with step-by-step explanations.
4/8/2026
Generalizable Audio-Visual Navigation via Binaural Difference Attention and Action Transition Prediction
BDATP framework for audio-visual navigation using binaural attention and action prediction to improve generalization in unseen 3D environments.
4/8/2026
Comparative Characterization of KV Cache Management Strategies for LLM Inference
Comparative analysis of key-value cache management strategies for efficient LLM inference under different model sizes and context lengths.
4/8/2026
Scaling Coding Agents via Atomic Skills
Proposes training LLM coding agents on five atomic coding skills (localization, editing, testing, reproduction, review) for improved generalization.
4/8/2026
StarVLA: A Lego-like Codebase for Vision-Language-Action Model Developing
StarVLA provides a modular open-source codebase for building vision-language-action embodied agents with standardized evaluation protocols.
4/8/2026
Phase-Associative Memory: Sequence Modeling in Complex Hilbert Space
Phase-Associative Memory is a recurrent sequence model using complex-valued representations achieving competitive perplexity on WikiText-103.
4/8/2026
PCA-Driven Adaptive Sensor Triage for Edge AI Inference
PCA-Triage is a streaming algorithm for adaptive sensor sampling in IoT networks using principal component analysis to manage bandwidth constraints.
4/8/2026
This Treatment Works, Right? Evaluating LLM Sensitivity to Patient Question Framing in Medical QA
Study evaluating LLM sensitivity to prompt phrasing in medical question answering, showing inconsistent responses despite identical underlying evidence.
4/8/2026
AutoLALA: Automatic Loop Algebraic Locality Analysis for AI and HPC Kernels
arXiv paper presenting AutoLALA, open-source tool analyzing data locality in loop programs for HPC and AI workloads.
4/8/2026
Nidus: Externalized Reasoning for AI-Assisted Engineering
arXiv paper on Nidus, a governance runtime using Claude, Gemini, Codex to mechanize V-model for AI-assisted software delivery.
4/8/2026
Beyond LLM-as-a-Judge: Deterministic Metrics for Multilingual Generative Text Evaluation
arXiv paper proposing OmniScore, deterministic evaluation metrics for multilingual text generation as alternative to LLM judges.
4/8/2026
Edit, But Verify: An Empirical Audit of Instructed Code-Editing Benchmarks
arXiv paper auditing code-editing benchmarks for LLMs, finding flaws in existing evaluation methods for instructed code modification.
4/8/2026
Vintix II: Decision Pre-Trained Transformer is a Scalable In-Context Reinforcement Learner
arXiv paper on Decision Pre-Trained Transformer for in-context reinforcement learning, enabling scalable generalist agent training.
4/8/2026
CRAB: Codebook Rebalancing for Bias Mitigation in Generative Recommendation
arXiv paper on CRAB method for mitigating popularity bias in generative recommendation systems via codebook rebalancing.
4/8/2026
$\pi^2$: Structure-Originated Reasoning Data Improves Long-Context Reasoning Ability of Large Language Models
arXiv paper presenting π² pipeline for curating reasoning data from structured sources to improve LLM long-context reasoning.
4/8/2026
Watch Before You Answer: Learning from Visually Grounded Post-Training
arXiv paper on vision-language models learning from grounded video data, finding text-only bias in video benchmarks.
4/8/2026
Offline RL for Adaptive Policy Retrieval in Prior Authorization
arXiv paper modeling prior authorization policy retrieval as MDP for adaptive decision-making in healthcare insurance.
4/8/2026
Reasoning Through Chess: How Reasoning Evolves from Data Through Fine-Tuning and Reinforcement Learning
arXiv paper on how reasoning evolves in language models through fine-tuning and RL, studied via chess task performance.