datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
thinking-steering-vectorssteering-vectors-llama70bpoolbench-steering-vectorscot-controllability-steering-vectors
CoT-controllability steering vectors — artifacts
Artifacts for the project "A 2,880-number steering vector gives a reasoning model the
chain-of-thought control that fine-tuning does" on gpt-oss-20b. Code + the master notebook +
generate_figures.py that load these artifacts:
https://github.com/redwoodresearch/automated-research-projects (folder
cot-controllability-steering-vectors).
Contents
steering_vectors/ — the headline frozen-weights steering vector… See the full description on the dataset page: https://huggingface.co/datasets/automated-alignment-science/cot-controllability-steering-vectors.Qwen3-0.6B-pts-steering-vectors
PTS Steering Vectors Dataset
A dataset of activation-based steering vectors created using the Pivotal Token Search (PTS) technique.
Details
Source: Generated using the PTS tool
Model: Qwen/Qwen3-0.6B
Dataset Structure
This dataset contains:
steering_vectors.jsonl: The main file with token-level steering vectors
Usage
These steering vectors can be used for activation-based steering during inference to guide language models toward particular… See the full description on the dataset page: https://huggingface.co/datasets/codelion/Qwen3-0.6B-pts-steering-vectors.cot-controllability-steering-vectors
CoT-controllability steering vectors — artifacts
Artifacts for the project "Activation steering can increase chain-of-thought
controllability" on gpt-oss-20b: a single frozen-weights steering vector (2,880 numbers added
to one layer's residual stream) matches what a LoRA fine-tune does to the model's CoT
controllability on held-out instructions, and works by raising the late attention heads' attention
onto the in-context instruction. Code + the master notebook +… See the full description on the dataset page: https://huggingface.co/datasets/ejcgan/cot-controllability-steering-vectors.DeepSeek-R1-Distill-Qwen-1.5B-pts-steering-vectors
PTS Steering Vectors Dataset
A dataset of activation-based steering vectors created using the Pivotal Token Search (PTS) technique.
Details
Source: Generated using the PTS tool
Model: deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B
Dataset Structure
This dataset contains:
steering_vectors.jsonl: The main file with token-level steering vectors
Usage
These steering vectors can be used for activation-based steering during inference to guide language… See the full description on the dataset page: https://huggingface.co/datasets/codelion/DeepSeek-R1-Distill-Qwen-1.5B-pts-steering-vectors.steering-vectors-openended-llama3.2-1B-itsteering-vectors-caa-phi3.5-mini-itmo-steering-vectors
Model-Organism Steering Vectors
Activation-steering direction tensors for the evaluation-aware model organisms
(Qwen3-32B, Llama-3.3-70B, Kimi-K2-Thinking).
Format
Each direction is a .pt file, shape [n_layers, hidden_size] (float32 or
float64). Index as direction[layer] to get the [hidden_size] vector for
a specific layer.
import torch
d = torch.load("qwen3-32b/eval_awareness/direction.pt", weights_only=True)
# d.shape == (64, 5120)
# d[28] == direction vector for layer… See the full description on the dataset page: https://huggingface.co/datasets/obalcells/mo-steering-vectors.steering-vectors-openended-qwen2.5-7b-itsteering-vectors-openended-phi3.5-mini-instructsteering-vectorssteering-vectors-openended-qwen2.5-7b-itsteering-vectors-openended-phi3.5-mini-instructsteering-vectors-openended-qwen2.5-7bsteering-vectors-openended-llama3.1-8B-itsteering-vectors-openended-self_awareness_general-qwen2.5-7b-itsteering-vectors-openended-phi3.5-mini-itsteering-vectors-openended-qwensteering-vectors-openended-llama3.1-8bintrospection-steering-vectorssteering-vectors-openended