datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
InfoSeek_emb_qwen3vle_2bcircuitlens-gemma-2-2btranscoder-descriptions-and-evaluations
CircuitLens & WeightLens: Transcoder Descriptions and Evaluations
This dataset contains automatically generated descriptions and evaluation metrics for Gemma-2-2B transcoders, produced using CircuitLens and WeightLens methods.
Methods
CircuitLens: https://github.com/egolimblevskaia/CircuitLens
WeightLens: https://github.com/egolimblevskaia/WeightLens
Dataset Structure
The dataset is organized by layers (0, 4, 7, 10, 12, 15, 18, 21, 23, 25), with each layer… See the full description on the dataset page: https://huggingface.co/datasets/egolimblevskaia/circuitlens-gemma-2-2btranscoder-descriptions-and-evaluations.diffing-stats-gemma-2-2b-crosscoder-l13-mu4.1e-02-lr1e-04 Contains maximum activating examples for all the features of our crosscoder trained on gemma 2 2B layer 13 available here: https://huggingface.co/Butanium/gemma-2-2b-crosscoder-l13-mu4.1e-02-lr1e-04/blob/main/README.md
base_examples.pt contains all the maximum examples of the feature on a subset of validation test of fineweb
chat_examples.pt is the same but for lmsys chat data
chat_base_examples.pt is a merge of the two above files.
All files are of the type dict[int, list[tuple[float… See the full description on the dataset page: https://huggingface.co/datasets/science-of-finetuning/diffing-stats-gemma-2-2b-crosscoder-l13-mu4.1e-02-lr1e-04.MindBigData2023_MNIST-2B
Dataset Summary
MindBigData 2023 MNIST-2B is a reduced subset of the MindBigData 2023 MNIST-8B https://huggingface.co/datasets/DavidVivancos/MindBigData2023_MNIST-8B (June 1st 2023), brain signals open dataset created for Machine Learning, based on EEG signals from a single subject captured using a custom 128 channels device, replicating the full 70,000 digits from Yaan LeCun et all MNIST dataset. The brain signals were captured while the subject was watching the pixels of the… See the full description on the dataset page: https://huggingface.co/datasets/DavidVivancos/MindBigData2023_MNIST-2B.capterra-b2b-software-reviews
Capterra B2B Software Reviews
56,606 B2B software reviews from Capterra, covering 66 products across 11 software categories.
Most public review datasets are star rating + review text. This one carries five separate rating dimensions, pros and cons as distinct pre-split fields, reviewer firmographics, and, unusually, an incentive disclosure flag recording whether the reviewer was given a gift card, referred by the vendor, or wrote the review unprompted.
Why this is… See the full description on the dataset page: https://huggingface.co/datasets/FreshCrawl/capterra-b2b-software-reviews.B2B_Sales_dataused in these articles:
https://www.sciencedirect.com/science/article/abs/pii/S0957417416306327
https://www.emerald.com/insight/content/doi/10.1108/imds-09-2016-0409/full/html
b2b-digital-marketing-performance-benchmarks
B2B & Ecommerce Performance Marketing Benchmarks
Maintained and published by Datametrik — Performance Marketing and Growth Agency.
diffing-stats-SAE-difference_cb-gemma-2-2b-L13-k100-x8-lr1e-04-local-shufflingsynthetic_polistance
Fully Synthetic Prompts for LLM Political Stance Detection
All resources developed in the article "Templated or fully Synthetic? Prompt construction as a confound in measuring LLM political stance beyond writing assistance" (Chalkidis, 2026).
Paper Abstract
Political stance detection in LLMs has long been dominated by closed-ended, multiple-choice political survey questions—originally designed for humans, and thus lacks the realism and nuance of human-AI… See the full description on the dataset page: https://huggingface.co/datasets/kiddothe2b/synthetic_polistance.ai-roi-b2b-france-200-deployments
AI ROI Dataset — 200 B2B AI deployments in France (2022-2025)
Author : Denis Atlan (ENDKOO) — ORCID 0009-0007-0785-7305
License : CC BY 4.0 · DOI : 10.5281/zenodo.17795133 · Version : 2.0 (July 2026)
Self-published observational dataset. Not peer-reviewed. No independent audit. The author was involved as a consultant or service provider in a majority of the documented deployments — see Declared limitations.
Headline figures
The central variable is a… See the full description on the dataset page: https://huggingface.co/datasets/ENDKOO/ai-roi-b2b-france-200-deployments.gemma_2b_outputs
Gemma 2B Green LLM Experiment Outputs
This dataset repository contains experiment artifacts for Gemma 2B green-LLM runs, including LoRA adapter checkpoints, metrics, predictions, carbon logs, and figures.
Contents
checkpoints/: LoRA adapter checkpoints for CE baseline and joint-loss variants.
metrics/: training histories, SQuAD and MMLU summaries, prediction CSVs, calibration tables, and surrogate weights.
logs/: run histories and carbon summary JSON files.
carbon/:… See the full description on the dataset page: https://huggingface.co/datasets/PhotonTJ/gemma_2b_outputs.angry-sir-ca0c2b
angry-sir-ca0c2b
Synthetic sensors test data: 42 rows in data.csv.
All values are randomly generated fictional examples, not real observations, products, or user activity. Intended only for CSV loading and pipeline tests; not suitable for scientific or business conclusions. Columns are sampled independently and do not model real-world correlations.
Fields
sample_id: random identifier for this generated sample.
row_id: sequential row number starting at 1.… See the full description on the dataset page: https://huggingface.co/datasets/hanakotanaka/angry-sir-ca0c2b.max-activating-examples-gemma-2-2b-l13-ckissanediffing-stats-SAE-base-gemma-2-2b-L13-k100-x32-lr1e-04-local-shufflingtitan-signal-b2b-ai-leads
Titan Signal - B2B AI Company Lead Intelligence
Verified contact records for decision-makers at AI, ML, and enterprise software companies.
Built by Titan Signal's 230+ autonomous harvesting agents. Continuously refreshed, MX-verified, 90-day auto-purge.
Fields
Company name, contact title, email
Industry vertical, headcount band, revenue band
Tech stack tags, engagement score, verification date
Full Dataset
This is a 50-record sample. Full datasets (10K-1M+… See the full description on the dataset page: https://huggingface.co/datasets/SophieTitan/titan-signal-b2b-ai-leads.diffing-stats-gemma-2-2b-L13-k100-lr1e-04-local-shuffling-CCLossdiffing-stats-SAE-difference_cb-gemma-2-2b-L13-k100-lr1e-04-local-shufflingdiffing-stats-SAE-difference_cb-gemma-2-2b-L13-k100-x2-lr1e-04-local-shufflingdiffing-stats-SAE-chat-gemma-2-2b-L13-k100-lr1e-04-local-shufflingd369-quran-fingerprint
d369 — Quranic {3,6,9} Digital Root Fingerprint
Author: Emad Suleiman Alwan | up2b.ai | ORCID: 0009-0004-5797-6140
License: CC BY 4.0 | Published: March 2026
Overview
This dataset accompanies a five-paper series documenting a statistically significant
numerical fingerprint in the Quran under the Special-6 (KHASS_6) encoding system.
Core finding: 51 of 114 Surahs (44.7%) have a digit root falling in {3, 6, 9}
under Special-6 encoding — a proportion that occurs by chance… See the full description on the dataset page: https://huggingface.co/datasets/up2b/d369-quran-fingerprint.diffing-stats-SAE-difference-gemma-2-2b-L13-k100-lr1e-04-local-shufflingyoutu-llm-2b-base-blind-spots
Youtu-LLM-2B-Base Blind Spots Dataset
What is this?
I tested a small AI language model called Youtu-LLM-2B-Base (made by Tencent) to find places where it gives wrong or strange answers. I gave it 50 different questions and kept the 25 cases where it clearly failed.
This dataset contains those 25 failures — the question I asked, what the correct answer should be, what the model actually said, and why it was wrong.
About the Model
Name: Youtu-LLM-2B-Base
Link:… See the full description on the dataset page: https://huggingface.co/datasets/Afras/youtu-llm-2b-base-blind-spots.diffing-stats-SAE-difference_cb-gemma-2-2b-L13-k100-x1-lr1e-04-local-shuffling2_big_en_1diffing-stats-gemma-2-2b-L13-k100-lr1e-04-local-shuffling-Decoupledgemma-2-2b-it-ipg-training-datadiffing-stats-gemma-2-2b-it-Meditron3-L16-k100-lr1e-04-local-shuffling-CCLossgemma-2b_self_0.7_0.05_wiki_sentencesdiffing-stats-gemma-2-2b-it-Meditron3-L16-mu3.8e-02-lr1e-04-local-shuffling-CCLossgemma-2b-cameroon-cultural-blindspots
Gemma-2b Cameroon Cultural Blindspots
This dataset highlights the "blind spots" of the Google Gemma-2-2b base model regarding Cameroonian culture, geography, and local languages.
1. Model Tested
Model Name: google/gemma-2-2b
Type: Base Model (Pre-trained)
2. Loading Procedure
The model was loaded using the transformers library on a Google Colab T4 GPU:
from transformers import AutoTokenizer, AutoModelForCausalLM
import torch
model_id = "google/gemma-2-2b"… See the full description on the dataset page: https://huggingface.co/datasets/zox-BT/gemma-2b-cameroon-cultural-blindspots.