datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
gspc-axis-corpus
GSPC axis transcripts — prompts and full responses
The transcript corpus behind graded runs. Each row of gspc-axis.jsonl carries the
axis, the model, the prompt it was given and the full response it returned. This is the raw material
a grader reads — publishing it is what lets a stranger re-grade our runs instead of trusting our grades.
The live board is the authority
GET https://councilof.ai/api/gspc — quote totals.public_count. This Hub card is a printer of… See the full description on the dataset page: https://huggingface.co/datasets/csoai/gspc-axis-corpus.assistant-axis-vectors
Assistant Axis Vectors for gemma-3-27b-it
This dataset contains pre-computed role vectors and the assistant axis for gemma-3-27b-it.
Overview
These vectors were computed using the methodology from the paper "The Assistant Axis"
by Christina Lu et al. The vectors can be used for activation steering to control model behavior along the
"assistant-like" to "role-playing" spectrum.
Contents
gemma-3-27b-it/assistant_axis.pt - The computed assistant axis (principal… See the full description on the dataset page: https://huggingface.co/datasets/massines3a/assistant-axis-vectors.arena-elo-by-axis
Arena Elo by axis (not a board score)
A pairwise-arena Elo reference computed per axis. elo_reference.json carries the
axis list, a content_id over the payload, and a leaderboard row per model: elo, games, winrate and a
confidence interval. This is an arena number over head-to-head rounds — it is not a GSPC board score and
does not write to any slot.
The live board is the authority
GET https://councilof.ai/api/gspc — quote totals.public_count. This Hub card is… See the full description on the dataset page: https://huggingface.co/datasets/csoai/arena-elo-by-axis.Test-Dataset
AXIS held-out 20 — a cross-embodiment few-shot adaptation benchmark
20 tasks never seen in pretraining, 20 demonstrations each, LoRA adaptation, rollout evaluation.
This release is the SPECIFICATION and the INDEX, not the demonstration data. It is published
first and on purpose: everything here is what you need to render the benchmark on your own
embodiment, and none of it depends on our video encoding being finished.
Status: the task list is CANDIDATES. The learnability gate… See the full description on the dataset page: https://huggingface.co/datasets/axisrobotics/Test-Dataset.dual-axis
Dual-Axis InteractionSpeech Dataset
This release accompanies Dual-Axis Generative Reward Model Toward Semantic and Turn-taking Robustness in Interactive Spoken Dialogue Models. It contains interaction-quality supervision derived from the English InteractionSpeech corpus and separates two evaluation dimensions:
response relevance: whether the response is semantically appropriate and consistent with the dialogue state;
interactional fluency: whether turn timing and floor… See the full description on the dataset page: https://huggingface.co/datasets/1f/dual-axis.Axis-1
🚀 Welcome to Axis-1: The Frontier of Machine Reasoning
Ready to push your LLMs past the superficial? Axis-1 isn't just another dataset—it’s a rigorous, high-fidelity playground designed for models that don't just guess, but think. 🧠
We’ve curated a collection of the most demanding problems in physics, mathematics, and algorithmic engineering to force models to abandon intuition-shortcuts and embrace structured, self-correcting logical paths.
🎯 What is Axis-1?… See the full description on the dataset page: https://huggingface.co/datasets/Axiom-AI/Axis-1.HEC3R-ckpt-fixed_view_axis_freeze_decarch-poison-mass-lpi-260903T0800-axisA-promptsaxis-chain
AXIS-Chain: Agentic Cross-layer Injection Chains
AXIS-Chain is a cross-layer prompt-injection benchmark for agentic systems. Public
evaluations test one layer in isolation (tool injection, or payments, or tool poisoning).
AXIS-Chain instead composes attack chains that enter at one protocol layer and are
scored at another, the final signed payment, and asks whether a defense stops the
end-to-end compromise rather than a single step.
This repository holds the data in two subsets… See the full description on the dataset page: https://huggingface.co/datasets/anonymos-2321135/axis-chain.chart-qa-with-axis-tricks-v1
This dataset is a remastered version prepared using Adaption's Adaptive Data platform.
chart_qa_with_axis_tricks
This dataset contains pairs of chart descriptions and questions focused on interpreting time-series data for meat production and forest area across various countries. The samples specifically highlight misleading visualization techniques, such as inverted axes, truncated y-axes, and non-zero baselines, requiring the model to calculate actual values and percentage… See the full description on the dataset page: https://huggingface.co/datasets/rodriguescarson/chart-qa-with-axis-tricks-v1.chart-qa-with-axis-tricks
This dataset is a remastered version prepared using Adaption's Adaptive Data platform.
chart_qa_with_axis_tricks
This dataset contains pairs of chart descriptions and questions focused on interpreting time-series data for meat production and forest area across various countries. The samples specifically highlight misleading visualization techniques, such as inverted axes, truncated y-axes, and non-zero baselines, requiring the model to calculate actual values and percentage… See the full description on the dataset page: https://huggingface.co/datasets/rodriguescarson/chart-qa-with-axis-tricks.adaption-chart-qa-with-axis-tricksThis dataset is a remastered version prepared using Adaption's Adaptive Data platform.
adaption-chart_qa_with_axis_tricks
This dataset contains pairs of chart descriptions and questions focused on interpreting time-series data for meat production and forest area across various countries. The samples specifically highlight misleading visualization techniques, such as inverted axes, truncated y-axes, and non-zero baselines, requiring the model to calculate actual values and… See the full description on the dataset page: https://huggingface.co/datasets/rodriguescarson/adaption-chart-qa-with-axis-tricks.beyond-single-axis-artifactsassistant-axis-unknown-self-chatassistant-axis-self-chat