GPQA-Diamond

The 'GPQA-Diamond' dataset/benchmark contains reasoning tasks used to evaluate the performance of quantized Large Reasoning Models (LRMs) during fine-tuning.

Papers using GPQA-Diamond (23)

Trust but Verify: Prover-Verifier Deliberation for Selective LLM Prediction2026Self-Verified Distillation: Your Language Model Is Secretly Its Own Synthetic Data Pipeline2026Transformation-Augmented GRPO for Enhancing Exploration in Reasoning of Large Language Models2026QuantLRM: Quantization of Large Reasoning Models via Fine-Tuning Signals2026Improving Data and Reward Design for Scientific Reasoning in Large Language Models2026Darwin Family: MRI-Trust-Weighted Evolutionary Merging for Training-Free Scaling of Language-Model Reasoning2026PRISM: Demystifying Retention and Interaction in Mid-Training2026Lie to Me: How Faithful Is Chain-of-Thought Reasoning in Reasoning Models?2026D-COT: Disciplined Chain-of-Thought Learning for Efficient Reasoning in Small Language Models2026Asking LLMs to Verify First is Almost Free Lunch2025Prompting Test-Time Scaling Is A Strong LLM Reasoning Data Augmentation2025Think Twice: Enhancing LLM Reasoning by Scaling Multi-round Test-time Thinking2025DianJin-R1: Evaluating and Enhancing Financial Reasoning in Large Language Models2025Process Reward Models That Think2025Prior Prompt Engineering for Reinforcement Fine-Tuning2025Void in Language Models2025First Finish Search: Efficient Test-Time Scaling in Large Language Models2025LIMOPro: Reasoning Refinement for Efficient and Effective Test-time Scaling2025Ring-lite: Scalable Reasoning via C3PO-Stabilized Reinforcement Learning for LLMs2025ReasonFlux-PRM: Trajectory-Aware PRMs for Long Chain-of-Thought Reasoning in LLMs2025Answer Matching Outperforms Multiple Choice for Language Model Evaluation2025Agentar-Fin-R1: Enhancing Financial Intelligence through Domain Expertise, Training Efficiency, and Advanced Reasoning2025Meta-Awareness Enhances Reasoning Models: Self-Alignment Reinforcement Learning2025