…On three heterogeneous task streams with Qwen3-8B, EVOCHAMBER reaches 63.9% on competition math, 75.7% on code, and 87.1% on multi-domain reasoning, outperforming the best baseline by 32…
…Our approach consistently improves LLM performance across multiple tasks, including instruction following and math reasoning, in both training free and fine tuning settings. Moreover, we show that our method mitigates attention sinks…
…Process-Driven Image Generation via Interleaved Reasoning (2026) Affordance Agent Harness: Verification-Gated Skill Orchestration (2026) MathGen: Revealing the Illusion of Mathematical Competence through Text-to-Image Generation (2026) Enhanced Text-to…
…AI-generated summary On-Policy Distillation (OPD) has emerged as a dominant post-training paradigm for large language models, especially for reasoning domains. However, OPD remains unstable in practice due to the…
…However, accurately interpreting unconstrained STEM student handwritten solutions with intertwined mathematical formulas, diagrams, and textual reasoning poses a significant challenge due to the lack of authentic and domain-specific benchmarks. Additionally, current…
…Amirreza Mohseni , , , Abstract CurveBench presents a benchmark for hierarchical topological reasoning using visual inputs, demonstrating significant challenges in exact topology-aware visual reasoning even with advanced models. AI-generated summary We introduce…
…Experiments on mathematical reasoning benchmarks show that the discovered strategies improve the overall accuracy--cost tradeoff over strong manually designed baselines. The discovered strategies generalize to held-out benchmarks and model scales…
…AI-generated summary Patent examination is a complex, multi-stage process requiring both technical expertise and legal reasoning , increasingly challenged by rising application volumes. Prior benchmark s predominantly view patent examination as…
…current AI agents unable to solve effectively. Curated from 230 student-submitted candidates through rigorous expert review, the final task set spans 25+ professional domains, ranging from olympiad-level mathematics and linguistics…
…AI-generated summary Large Language Models often improve accuracy on reasoning tasks by sampling multiple Chain-of-Thought (CoT) traces and aggregating them with majority voting (MV), a test-time technique called…