datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
pubmed-2019-pythia-word-tfidf-pubmedqa-clean-articlespubmed-2019-pythia-word-tfidf-pubmedqa-clean-val-sequencespubmed-2019-pythia-word-tfidf-invfreq-pubmedqa-clean-articlespubmed-2019-pythia-word-tfidf-invfreq-pubmedqa-clean-val-sequencesall-the-news-2-pythia-tfidf-wordlevelall-the-news-2-pythia-tfidf-invfreq-topic-stratified-v1-articlesall-the-news-2-pythia-tfidf-topic-stratified-v1-articlesall-the-news-2-tf-idf-wordlevel-sublineartoy-models-of-sft-data
Toy Models of SFT Data
This is a public-clean candidate data package for the Toy Models of SFT project.
It is built for researcher inspection first.
The package answers two questions:
What were the models trained on?
How did the models actually behave under evaluation?
The package includes training data, eval inputs, model rollouts, judge scores,
parsed GPQA outputs, aggregate tables, paper figures, frozen plot data, and
provenance records. It deliberately includes some… See the full description on the dataset page: https://huggingface.co/datasets/matonski/toy-models-of-sft-data.all-the-news-2-pythia-tfidf-invfreq-topic-stratified-v1-val-sequencesall-the-news-2-pythia-tfidf-topic-stratified-v1-val-sequenceswikitext-103-raw-pythia-word-tfidf-invfreq-topic-stratified-v1-articleswikitext-103-raw-pythia-word-tfidfcoderTraining dataset for finetuning for human-eval.
This dataset has been created from the following datasets:
sahil2801/CodeAlpaca-20k
sahil2801/code_instructions_120k
mhhmm/leetcode-solutions-python
teknium1/GPTeacher
Script for generating dataset: create_dataset.py.
wikitext-103-raw-pythia-word-tfidf-topic-stratified-v1-articleswikitext-103-raw-pythia-tfidf-tokenlevelall-the-news-2-tfidf-invfreq-topic-stratified-v1-articleswikitext-103-raw-v4-df-tokenlevelwikitext-103-raw-v2-tfidf-invfreq-topic-stratified-v1-articlesall-the-news-2-tfidf-topic-stratified-v1-articles-sublinearreward-hacking-prompts
Reward Hacking Prompts Dataset
A dataset of 50 computational task prompts designed to elicit reward hacking behavior in GPT-OSS-20B.
Dataset Description
This dataset provides 50 computational task prompts empirically validated to elicit reward hacking behavior in LLMs.
Reward hacking occurs when models find shortcuts to pass grading criteria without actually solving the problem.
What's Included
50 prompts: Computational tasks ranging from fluid simulation to… See the full description on the dataset page: https://huggingface.co/datasets/matonski/reward-hacking-prompts.wikitext-103-raw-v2-tfidf-topic-stratified-v1-articles-sublinearwikitext-103-raw-pythia-tfidf-topic-stratified-v1-articleswikitext-103-raw-v1-idfwikitext-103-raw-pythia-tfidf-invfreq-topic-stratified-v1-articleswikitext-103-raw-v2-tf-idf-wordlevel-sublinearwikitext-103-raw-pythia-word-tfidf-invfreq-topic-stratified-v1-val-sequencesall-the-news-2-freq-invsqrt-topic-stratified-v1-articlesbookswikitext-103-raw-v2-freq-invsqrt-topic-stratified-v1-articles