datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
cot_control_controllability_2000cot_control_deepseek_v3_controllability_2000cot_control_kimi_k2_controllability_2000cot_control_qwen3_32b_controllability_2000fusion-hybrid-controllability-loss-horizon-and-control-routing-v0.1What this dataset tests
Forecasts when a hybrid reactor will lose controllable power.
Maps stability margin collapse into a time horizon.
Routes optimal intervention.
Required outputs
time_to_control_loss_min
recommended_action
Use case
Predictive stabilization for subcritical or hybrid reactors.
Prevents loss of controllability before shutdown conditions.
cot_control_kimi_k25_controllability_2000itts-controllability
CommonVoice ITTS Controllability Validation Set
Ground-truth-labelled English speech clips drawn from
fixie-ai/common_voice_17_0
(Common Voice 17.0, CC0), used to validate audio-LM judges for
instructable-TTS (ITTS) controllability scoring on age, gender, and
native accent.
Each clip carries a human-provided attribute label from Common Voice, so a judge's
perceived-attribute accuracy can be measured against real ground truth (blind
perceived-accuracy protocol).… See the full description on the dataset page: https://huggingface.co/datasets/Snooow1029/itts-controllability.cot_control_gpt_oss_20b_controllability_2000