datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
BLIP3o-NEXT-EDIT-ENSEMBLE-DATASETSllm-ensembles-20-afg-vs-ug-gradual-rhollm-ensembles-18-pilot-multilingual-mixllm-ensembles-26-narrow-kodigit_mask_ensemble_distilled_from_cv12_balanced_mfcc
Dataset Card for "digit_mask_ensemble_distilled_from_cv12_balanced_mfcc"
More Information needed
llm-ensembles-15-afg-m50-diverse0_digit_mask_ensemble_distilled_from_cv12_balanced_mfcc
Dataset Card for "0_digit_mask_ensemble_distilled_from_cv12_balanced_mfcc"
More Information needed
0-9up_ft_ensemble_distilled_mfcc
Dataset Card for "0-9up-ft_ensemble_distilled_mfcc"
More Information needed
1_digit_mask_ensemble_distilled_from_cv12_balanced_mfcc
Dataset Card for "1_digit_mask_ensemble_distilled_from_cv12_balanced_mfcc"
More Information needed
Phi4-ensemble-teacher-forcing-record-logits-datallm-ensembles-25-fix-poolllm-ensembles-13-inspect-evals-validationdigit_mask_ft_ensemble_distilled_mfcc
Dataset Card for "digit_mask_ft_ensemble_distilled_mfcc"
More Information needed
eval_ACT_RJ45_piper_20260113_ckpt40k_ensemble_50runsThis dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v2.1",
"robot_type": "piper",
"total_episodes": 50,
"total_frames": 20487,
"total_tasks": 1,
"total_videos": 100,
"total_chunks": 1,
"chunks_size": 1000,
"fps": 30,
"splits": {
"train": "0:50"},
"data_path": "data/chunk-{episode_chunk:03d}/episode_{episode_index:06d}.parquet",
"video_path":… See the full description on the dataset page: https://huggingface.co/datasets/BensoAI/eval_ACT_RJ45_piper_20260113_ckpt40k_ensemble_50runs.roadsign-judged-ensemble-agree1
Box-overlay preview
Auto-generated sample of the labelled boxes (with judge scores when available). Regenerated on every push.
details_TFLai__Ensemble5-Platypus2-13B-QLora-0.80-epoch
Dataset Card for Evaluation run of TFLai/Ensemble5-Platypus2-13B-QLora-0.80-epoch
Dataset Summary
Dataset automatically created during the evaluation run of model TFLai/Ensemble5-Platypus2-13B-QLora-0.80-epoch on the Open LLM Leaderboard.
The dataset is composed of 64 configuration, each one coresponding to one of the evaluated task.
The dataset has been created from 2 run(s). Each run can be found as a specific split in each configuration, the split being named using the… See the full description on the dataset page: https://huggingface.co/datasets/open-llm-leaderboard-old/details_TFLai__Ensemble5-Platypus2-13B-QLora-0.80-epoch.p2-llm-etf-ensemble-resultsopeninterp-45-inference-ensemble
nb45 — Inference-Time Multi-Probe Ensemble (FG + RG fusion)
Headline result: combining FabricationGuard + ReasonGuard at inference time — applied to base Qwen3.6-27B with no retraining — produces +6.7 percentage points AUROC over the best single probe in detecting incorrect answers, at 0.18 ms total inference overhead per generation (Path A, vLLM streaming) or 148 ms (Path B, post-generation capture).
This validates the ProbePack product angle: ship probe ensembles as middleware… See the full description on the dataset page: https://huggingface.co/datasets/caiovicentino1/openinterp-45-inference-ensemble.daily-paper-2026-08-06-ensemble-tier-voting-skill-router
Cross-Tier Model Ensemble Voting for Skill Routing: A Cost-Quality Study Beyond Single-Signal Retrieval Fusion
TL;DR — An analytical cost-quality framework for adding confidence-weighted model-tier voters to an existing free hybrid retrieval signal, with pre-registered evaluation protocol and break-even thresholds for four deployment arms.
ThakiCloud AI Research · 2026-08-06 · 📝 Tech blog (KO)
Problem
Production skill routers reduce routing decisions to a… See the full description on the dataset page: https://huggingface.co/datasets/thaki-AI/daily-paper-2026-08-06-ensemble-tier-voting-skill-router.0-9up_ft_ensemble_distilled_from_cv12_balanced_mfcc
Dataset Card for "0-9up_ft_ensemble_distilled_from_cv12_balanced_mfcc"
More Information needed
details_PulsarAI__EnsembleV5-Nova-13B
Dataset Card for Evaluation run of PulsarAI/EnsembleV5-Nova-13B
Dataset Summary
Dataset automatically created during the evaluation run of model PulsarAI/EnsembleV5-Nova-13B on the Open LLM Leaderboard.
The dataset is composed of 64 configuration, each one coresponding to one of the evaluated task.
The dataset has been created from 2 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The "train"… See the full description on the dataset page: https://huggingface.co/datasets/open-llm-leaderboard-old/details_PulsarAI__EnsembleV5-Nova-13B.docvqa-media-judged-ensemble
Box-overlay preview
Auto-generated sample of the labelled boxes (with judge scores when available). Regenerated on every push.
llm-classification-v7a-ensemble-teacher-v1
llm-classification-v7a-ensemble-teacher-v1
Pseudo-label dataset for Kaggle LLM Classification Finetuning v7a.
Purpose
This dataset is generated by 01.5 ensemble pseudo-labeling.
The purpose of this dataset is to transfer the knowledge of a two-model teacher ensemble into a single student model for v7a training.
Teachers:
v5a: tussiiiii/llmcmp-distill-llama3-8b-lora-v5a-no-rationale-long-ab-swap-merged
v6q:… See the full description on the dataset page: https://huggingface.co/datasets/tussiiiii/llm-classification-v7a-ensemble-teacher-v1.roadsign-judged-ensemble-agree2
Box-overlay preview
Auto-generated sample of the labelled boxes (with judge scores when available). Regenerated on every push.
docvqa-media3-judged-ensemble-v2-agree1
Box-overlay preview
Auto-generated sample of the labelled boxes (with judge scores when available). Regenerated on every push.
seismic-ensemble
Seismic Ensemble — Data & Model Store
This repo holds the data and model artefacts for the code at
irp-jas25, a project on multi-class seismic event discrimination (earthquake / deep earthquake / explosion /
nuclear explosion / volcanic eruption / noise) for CTBT-style monitoring, using a six-member ensemble
(four deep models, two classical) with a stacked meta-learner on top. It accompanies the paper
"Robust and Explainable Multi-Class Seismic Event Discrimination through… See the full description on the dataset page: https://huggingface.co/datasets/Jamie1701/seismic-ensemble.TOA-Ultrafeedback-SFT-Ensemble-model-num-4vgic-mutant-structural-ensembles
VGIC Mutant Structural Ensembles
Structural ensemble dataset generated for the manuscript:
Targeted MSA Masking Reshapes Structural Sampling in Mutant Voltage-Gated Ion Channels
Dataset summary
This repository contains predicted structural ensembles of voltage-gated ion channels generated for the study “Targeted MSA Masking Reshapes Structural Sampling in Mutant Voltage-Gated Ion Channels.”
The dataset was created to investigate how targeted masking… See the full description on the dataset page: https://huggingface.co/datasets/adrishgz/vgic-mutant-structural-ensembles.docvqa-media3-judged-ensemble-v2
Box-overlay preview
Auto-generated sample of the labelled boxes (with judge scores when available). Regenerated on every push.
sms-ensemble-fitting