Human Feedback and Eval Paper Explorer

Match reason: Keyword overlap 1/1 across title and protocol fields.

• When an LLM-based agent improves on a task, is the gain from the model itself or from the reasoning paradigm wrapped around it?
• We study this question by comparing six inference-time paradigms, namely Direct, CoT, ReAct, Plan-Execute, Reflection, and ReCode, across four frontier LLMs and ten benchmarks, yielding roughly 18,000 runs.

Match reason: Keyword overlap 1/1 across title and protocol fields.

• Experiments across a range of reasoning benchmarks including MATH500, AIME24/25, HumanEval, and MBPP show that our approach yields better exploration-quality tradeoff than both random and low-confidence remasking.

Match reason: Keyword overlap 1/1 across title and protocol fields.

• Using roughly 48 execution-verified HumanEval training solutions, tuning a single initial state matrix per recurrent layer, with zero inference overhead, outperforms LoRA by +10.8 pp (p < 0.001) on HumanEval.
• Cross-domain transfer is significant on MATH-500 (+4.8 pp, p = 0.00002, 8 seeds) and GSM8K (+2.8 pp, p = 0.0003, 10 seeds); a text-to-SQL benchmark (Spider) shows no transfer, consistent with the trajectory-steering mechanism.

Match reason: Keyword overlap 1/1 across title and protocol fields.

• To address this gap, we introduce CangjieBench, a contamination-free benchmark for Cangjie, a representative low-resource general-purpose language.
• Agent achieve state-of-the-art accuracy but incur high token consumption.

Match reason: Keyword overlap 1/1 across title and protocol fields.

• We evaluate our method on the HumanEval and GSM8K benchmarks using the LLaDA-8B-Instruct model.

Match reason: Keyword overlap 1/1 across title and protocol fields.

• Extensive experiments across seven benchmarks demonstrate that our ReflexiCoder-8B establishes a new state-of-the-art (SOTA) among leading open-source models in the 1.5B-14B range, achieving 94.51% (87.20%) on HumanEval (Plus), 81.80%…

Match reason: Keyword overlap 1/1 across title and protocol fields.

• Using controlled contamination experiments on GSM8K, HumanEval, and MATH, we find that CDD's effectiveness depends critically on whether fine-tuning produces verbatim memorization.

Match reason: Keyword overlap 1/1 across title and protocol fields.

• EngGPT2 is trained on 2.5 trillion tokens - less than Qwen3's 36T or Llama3's 15T - and delivers performance on key benchmarks, including MMLU-Pro, GSM8K, IFEval and HumanEval, comparable to dense models in the 8B-16B range, while requiring…

Match reason: Keyword overlap 1/1 across title and protocol fields.

• We introduce a taxonomy of three defense categories -- output perturbation, data poisoning, and information throttling -- and evaluate nine defense configurations using a standardized pipeline with Qwen3-14B as teacher and…

Match reason: Keyword overlap 1/1 across title and protocol fields.

• We evaluate our approach on four benchmarks with diverse tasks -- GSM8K (reasoning), HumanEval (code generation), IfEval (instruction following), and LongFormQA (question answering) -- revealing massive efficiency gains at minimal…

Match reason: Keyword overlap 1/1 across title and protocol fields.

• This suggests a simple intervention: by taking a Union over different extractors, we can increase the token yield of DCLM-Baseline by up to 71% while maintaining benchmark performance.

Match reason: Keyword overlap 1/1 across title and protocol fields.

• Across mathematical reasoning and code generation benchmarks (GSM8K, MBPP, HumanEval) on Dream-7B and LLaDA-8B, SOAR improves generation quality while maintaining competitive inference speed, offering a practical way to balance quality and…

Match reason: Keyword overlap 1/1 across title and protocol fields.

• Additional experiments show that the core principle of SLM-MUX extends to open-ended generation tasks (e.g., HumanEval) and benefits other model classes, including frontier LLMs and domain-specific fine-tuned SLMs.

Match reason: Keyword overlap 1/1 across title and protocol fields.

• Our analysis shows that there is a strong correlation between the model's single-sample accuracy and its false positive rate on HumanEval and MBPP, whose unit tests have limited coverage.

Match reason: Keyword overlap 1/1 across title and protocol fields.

• To address this, we introduce the Dynamic Auction-based Language Agent (DALA), a novel framework that treats communication bandwidth as a scarce and tradable resource.
• Extensive and comprehensive experiments demonstrate that our economically-driven DALA achieves new state-of-the-art performance across seven challenging reasoning benchmarks, including 84.32% on MMLU and a 91.21% pass@1 rate on HumanEval.