Poster Sessions

The accepted papers are split into two poster sessions in Hall A (Board 100-119). Each paper is presented in one session only.

  • Session 1 — 13:00–14:30
  • Session 2 — 15:20–16:50

Please find your assigned session below.

Session Paper
2 Adversarial Review: Structured Disagreement for Grounded Agentic Code Review
1 Agentic Neural Architecture Search
2 AI Coding Benchmarks Need Proofs, Not Just Tests
2 An Empirical Study of Proactive Coding Assistants in Real-World Software Development
1 ATLAS: Automated Toolkit for Large-Scale Verified Code Synthesis
2 Back to the Beginning of Heuristic Design: Bridging Code and Knowledge with LLMs
2 Benchmarks as Software: A Case Study on Terminal-Bench
2 Beyond Isolated Tasks: A Framework for Evaluating Coding Agents on Sequential Software Evolution
1 Beyond Lexical Similarity: A Benchmark for Evaluating Code Documentation Agents
1 Beyond Pass@k: Developing Human-Centric Evaluation Metrics for AI Pair Programmers
2 Beyond Problem Solving: UOJ-Bench for Evaluating Code Generation, Hacking, and Repair in Competitive Programming
1 BOAD: Discovering Hierarchical Software Engineering Agents via Bandit Optimization
1 Bridging Online and Offline RL: Contextual Bandit Learning for Multi-Turn Code Generation
2 Can LLMs Detect Benchmark Defects? A Meta-Benchmark from Benchmark Updates
2 Certifying the Judge: Falsifiable Properties for LLM-Based Evaluation of Formal Code
1 Coding With “Enemy”: Can Human Developers Detect AI Agent Sabotage?
1 Comparing Developer and LLM Biases in Code Evaluation
2 CRCA-Context: Counterfactual Robustness of Repository Context Retrieval Under Equivalent Issue Descriptions
1 DeRL-SWE: Decoupled Reinforcement Learning with Nested Credit Assignment for Software Engineering
1 Detecting Functional Memorization in Code Language Models
2 DevBench: An Interaction-Grounded Benchmark for Code Completion Models
1 Do AI Agents Write Less Maintainable Code Than Human Developers?
1 Domain-Adaptable Reinforcement Learning for Code Generation with Dense Rewards
2 Don’t Let Gains FADE: Breaking Down Policy Gradient Weights in RL
2 Don’t Claim Benchmark-Oriented Optimization Improves General Coding Capability — Diverse Evaluation Is Required
1 Evolutionary Multi-Task Optimization for LLM-Guided Discovery
2 Execution-Grounded Agents: Enforcing Physical Constraints in AI Code Generation via Oracle Search
1 Extracting Recurring Vulnerabilities from Black-Box LLM-Generated Software
2 ExVerus: Verus Proof Repair via Counterexample Reasoning
2 From Search to Policy: Thompson Sampling Trees for Robust LLM Code Refinement
1 Frozen Inner Looping Is Brittle for Code Generation with BitNet b1.58
1 GameDevBench: Evaluating Agentic Capabilities Through Game Development
1 Hawkeye: Hardware-Aware GPU Kernel Optimization with Minimal Supervision
2 Hidden Positives: Why Code Retrieval Benchmarks Underestimate Model Quality
1 How can we assess human-agent interactions? Case studies in software agent design
2 How Should Agents Read Demonstrations? Hierarchical Structure Beats Flat Action Logs
2 Improving Code Efficiency with Iterative Refinement using LLMs
1 Interactive Benchmarking of Scientific Coding Agents for Spatial Transcriptomics Alignment
2 Interpreting Code Correctness in Language Models through Activation Steering
1 JAXBench: Benchmarking Autonomous TPU Kernel Optimization
2 kAgent: An execution-guided crash resolution agent for the Linux kernel
1 Kernel-Smith: A Unified Recipe for Evolutionary Kernel Optimization
2 KForge: A Multi-Agent System for Cross-Platform Kernel Synthesis
1 Language-based Trial and Error Falls Behind in the Era of Experience
2 Learning Bug Context for PyTorch-to-JAX Translation with LLMs
1 LibEvoBench: Probing Temporal Knowledge Stratification in Code Generation Models
2 LLM-Based Rust Code Generation with On-the-Fly Compiler Feedback
1 Loop or Leap? Benchmarking Iterative vs Recursive Reasoning in Code LLMs with LRLBench
1 Making Execution Time a Trainable Reward for Code Generation
1 Matching Decompilation as a Verifier-Guided Task for Human-Centered Coding Agents
1 Metal-Sci: A Scientific Compute Benchmark for Evolutionary LLM Kernel Search on Apple Silicon
1 MultiVulnBench: A Large-Scale Benchmark for Count Bias in LLM-Based Multi-Vulnerability Detection
1 Nexus: Execution-Grounded Multi-Agent Test Oracle Synthesis
2 NKI-Agent: Domain-Specific Fine-Tuning and Agentic Tool Use for Neuron Kernel Generation
2 On Data Engineering for Scaling LLM Terminal Capabilities
2 optimize_anything: Unified Text Optimization can Outperform Specialized Systems
1 Orchestrating LLMs as Hierarchical Multi-Agent Reinforcement Learning System for Automotive Software Development
1 ParallelKernelBench: Can LLMs Write Fast Multi-GPU Kernels?
1 ProgramBench: Can Language Models Rebuild Programs From Scratch?
2 QuantumSemEval: Benchmarking LLMs’ Understanding of Quantum Programs via Semantic Equivalence Checking
2 Rapid Fixes, Gradual Failures: Exploring Iterative Self-Correction Dynamics in Large Language Models for Program Synthesis
1 RepoGuardBench: Repository-Borne Prompt Injection Attacks and Lightweight Defenses for Local Coding Agents
1 Safety Drift in Human-Centered Coding Agents: Suppression vs. Representation Shaping After Benign Adaptation
2 Scalable and Transparent Attribution for Human-AI Collaborative Code
2 Selective Code Generation under Correctness and Security Risk
2 Self-Distillation Zero: Self-Revision Turns Binary Rewards into Dense Supervision
1 SFT Overtraining Predicts Rank Inversion via Entropy Collapse Under RLVR
1 SkillFlow: Scalable and Efficient Agent Skill Retrieval System
2 SLEEP: Simulated Future Learning Environments for Automated Evolution of Heuristic Portfolios
2 SpIDER: Spatially Informed Dense Embedding Retrieval for Software Issue Localization
2 Steerability via constraints: a substrate for scalable oversight of coding agents
1 Step Rejection Fine-Tuning: A Practical Distillation Recipe
1 SWE Atlas: Benchmarking Coding Agents Beyond Issue Resolution
1 SWE-chat: Coding Agent Interactions From Real Users in the Wild
2 SWE-Router: Routing in Multi-turn Agentic Software Engineering Tasks
1 SynKer: Synthesize, Kernelize, Reinforce - Teaching GPU Kernel Generation to Small Language Models
1 Systematic LLM Translation of Legacy Scientific Code to Differentiable Frameworks: Application to a Land Surface Model
1 Teaching LLMs Program Semantics via Symbolic Execution Traces
2 Test-Time Scaling with Weak Verifiers via Self-Play
1 The Scaffold Effect in Coding Agents: Harness Choice as a Hidden Variable in Coding-Agent Evaluation
2 Towards Evaluation of Implicit Software World Models in Coding LLMs
1 TritonRL: Training LLMs to Think and Code Triton Without Cheating
1 Understanding and supporting how developers prompt for LLM-powered code editing in practice
2 V1: Unifying Generation and Self-Verification for Parallel Reasoners
2 VeriBench: An End-to-End Formal Verification Benchmark for AI Coding Agents in Lean 4
2 VibeServe: Can AI Agents Build Bespoke LLM Serving Systems?
2 VibeSWEBench: Can AI Co-Worker Agents Do Real-World Software Engineering by Vibe Coding?
2 Why Solve It Twice? Hierarchical Accumulation of Skills for Transfer-Efficient ML Engineering