datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
missing-value-function-interim-report
Between Hidden States and Control: Hidden-State Signals in Iterative LLM Repair
The Missing Value Function — Interim Research Report
Can minimal, non-learned signals derived from hidden states during inference serve as an internal value function to distinguish productive from unproductive revisitation in large language models?
Links
Resource
URL
📄 Zenodo (DOI, citable)
https://doi.org/10.5281/zenodo.18941566
📦 This repository… See the full description on the dataset page: https://huggingface.co/datasets/airVen/missing-value-function-interim-report.OpenR1-Stage0-DeepSeek-R1-Distill-Qwen-7B-Reasoning_Value_Functionvalue-function_eval_resultsSTAR-41K-DA-Stage0-DeepSeek-R1-Distill-Qwen-7B-Reasoning_Value_Function