tankiit/credence-ablation-roberta-n-heads-5-hatexplain
0
credence-ablation-roberta-n-heads-5-hatexplain
CREDENCE checkpoint for hatexplain (toxicity) with backbone roberta-base (5 concept heads).
Paper: https://huggingface.co/papers/2604.24170 · arXiv: 2604.24170
Training run folder: head_ablation_roberta_20251224_105619
Files
credence_checkpoint.pt— PyTorch checkpoint (model_state_dict,config,metadata, optimizer state)
Load with huggingface_hub
from huggingface_hub import hf_hub_download
import torch
path = hf_hub_download(repo_id="tankiit/credence-ablation-roberta-n-heads-5-hatexplain", filename="credence_checkpoint.pt")
ckpt = torch.load(path, map_location="cpu", weights_only=False)
state = ckpt["model_state_dict"]
config = ckpt["config"]Load into your CREDENCE model implementation (see project credence.py).
Citation
@article{mukherjee2026credence,
title={Credal Concept Bottleneck Models for Epistemic--Aleatoric Uncertainty Decomposition},
author={Mukherjee, et al.},
journal={arXiv preprint arXiv:2604.24170},
year={2026}
}