- Duel-Evolve: Reward-Free Test-Time Scaling via LLM Self-Preferences
Sweta Karlekar, Carolina Zheng, Magnus Saebo, Nicolas Beltran-Velez, Shuyang Yu · Feb 25, 2026 · Citations: 0
Pairwise Preference Automatic Metrics
Building on this observation, we introduce Duel-Evolve, an evolutionary optimization algorithm that replaces external scalar rewards with pairwise preferences elicited from the same LLM used to generate candidates.
- Surgical Post-Training: Cutting Errors, Keeping Knowledge
Wenye Lin, Kai Han · Mar 2, 2026 · Citations: 0
Pairwise Preference Automatic Metrics
While prior research emphasizes the role of on-policy data in mitigating forgetting, we uncover--and validate both theoretically and empirically--an overlooked yet critical mechanism: the implicit regularization inherent in Direct…
- Cache What Lasts: Token Retention for Memory-Bounded KV Cache in LLMs
Ngoc Bui, Shubham Sharma, Simran Lamba, Saumitra Mishra, Rex Ying · Dec 3, 2025 · Citations: 0
Automatic Metrics Long Horizon
Across mathematical reasoning (GSM8K, MATH-500, AIME24), procedural generation (LongProc), conversational long-memory benchmarks (LongMemEval), and long-context understanding (LongBenchV2 and SCBench), TRIM-KV consistently outperforms…
- BankMathBench: A Benchmark for Numerical Reasoning in Banking Scenarios
Yunseung Lee, Subin Kim, Youngjun Kwak, Jaegul Choo · Feb 19, 2026 · Citations: 0
Automatic Metrics Long Horizon
However, such errors have rarely been captured by existing benchmarks.
- Gradient Regularization Prevents Reward Hacking in Reinforcement Learning from Human Feedback and Verifiable Rewards
Johannes Ackermann, Michael Noukhovitch, Takashi Ishida, Masashi Sugiyama · Feb 20, 2026 · Citations: 0
Llm As JudgeAutomatic Metrics
Reinforcement Learning from Human Feedback (RLHF) or Verifiable Rewards (RLVR) are two key steps in the post-training of modern Language Models (LMs).
- GenDB: The Next Generation of Query Processing -- Synthesized, Not Engineered
Jiale Lao, Immanuel Trummer · Mar 2, 2026 · Citations: 0
Automatic Metrics Multi Agent
As a proof of concept, we present GenDB, an LLM-powered agentic system that generates instance-optimized and customized query execution code tailored to specific data, workloads, and hardware resources.
- Recycling Failures: Salvaging Exploration in RLVR via Fine-Grained Off-Policy Guidance
Yanwei Ren, Haotian Zhang, Likang Xiao, Xikai Zhang, Jiaxing Huang · Feb 27, 2026 · Citations: 0
Automatic Metrics Long Horizon
To address these issues, we propose SCOPE (Step-wise Correction for On-Policy Exploration), a novel framework that utilizes Process Reward Models to pinpoint the first erroneous step in suboptimal rollouts and applies fine-grained,…
- Test-Time Scaling with Diffusion Language Models via Reward-Guided Stitching
Roy Miles, Aysim Toker, Andreea-Maria Oncescu, Songcen Xu, Jiankang Deng · Feb 26, 2026 · Citations: 0
Automatic Metrics Long Horizon
This modular pipeline separates exploration (diffusion) from evaluation and solution synthesis, avoiding monolithic unified hybrids while preserving broad search.
- GATES: Self-Distillation under Privileged Context with Consensus Gating
Alex Stein, Furong Huang, Tom Goldstein · Feb 24, 2026 · Citations: 0
Automatic Metrics Long Horizon
Held-out in-domain accuracy under asymmetric evaluation improves from 46.0\% to 62.0\%, and average (maj@8) accuracy on public document-free math benchmarks improves from 20.2\% to 35.4\%.
- AgentDropoutV2: Optimizing Information Flow in Multi-Agent Systems via Test-Time Rectify-or-Reject Pruning
Yutong Wang, Siyuan Xiong, Xuebo Liu, Wenkang Zhou, Liang Ding · Feb 26, 2026 · Citations: 0
Automatic Metrics Multi Agent
We propose AgentDropoutV2, a test-time rectify-or-reject pruning framework designed to dynamically optimize MAS information flow without retraining.
- Orthogonalized Policy Optimization:Policy Optimization as Orthogonal Projection in Hilbert Space
Wang Zixian · Jan 18, 2026 · Citations: 0
Automatic Metrics Long Horizon
Experiments on MATH benchmarks show that the Hilbert projection formulation prevents gradient saturation typical of KL-constrained methods.
- Replaying pre-training data improves fine-tuning
Suhas Kotha, Percy Liang · Mar 5, 2026 · Citations: 0
Automatic Metrics Web Browsing
We demonstrate the success of replay in practice for fine-tuning 8B parameter models, improving agentic web navigation success by 4.5\% and Basque question-answering accuracy by 2\%.
- ToolMATH: A Math Tool Benchmark for Realistic Long-Horizon Multi-Tool Reasoning
Hyeonje Choi, Jeongsoo Lee, Hyojun Lee, Jay-Yoon Lee · Feb 24, 2026 · Citations: 0
Automatic Metrics Long Horizon
We introduce \ToolMATH, a math-grounded benchmark that evaluates tool-augmented language models in realistic multi-tool environments where the output depends on calling schema-specified tools and sustaining multi-step execution.
- Do LLMs and VLMs Share Neurons for Inference? Evidence and Mechanisms of Cross-Modal Transfer
Chenhang Cui, An Zhang, Yuxin Chen, Gelei Deng, Jingnan Zheng · Feb 22, 2026 · Citations: 0
Automatic Metrics Long Horizon
Across diverse mathematics and perception benchmarks, SNRF consistently enhances LVLM inference performance while preserving perceptual capabilities.
