Running
CATT token attribution demo
🔍
Ask a fine‑tuned LLM and see which documents taught it
CATT
community
AI & ML interests
None defined yet.
Recent Activity
Organization Card
CATT: Controlled Attribution Benchmark
Models, attribution scores, and demo for Elucidating the Design Space of LM Data Attribution (Steven Yu, Sheng-Yu Wang, Kimia Hamidieh, Antonio Torralba, David Bau).
Project Page | Demo | Code
Models
| Task | Base model (added tokens, not fine-tuned) | Fine-tuned LoRA adapters |
|---|---|---|
| Function Bindings | Qwen3.5-4B-d0-vtok101-base | Qwen3.5-4B-d0-vtok101-lora-seeds |
| Function Bindings with distractors | Qwen3.5-4B-d0-vtok101-distr-base | Qwen3.5-4B-d0-vtok101-distr-lora-seeds |
| Rule Routings | Qwen/Qwen3.5-4B | Qwen3.5-4B-route-ab-l8-lora-scale |
Attribution scores
Every method's scores on every model in the sweep, with their recall and AUC metrics: Function Bindings, Function Bindings with distractors, Rule Routings.
FTRACE-TREX
The remove-and-retrain experiment on FTRACE-TREX: retraining results, attribution scores, and mT5-small models.
models 6
catt-bench/Qwen3.5-4B-d0-vtok101-distr-lora-seeds
Updated • 2
catt-bench/Qwen3.5-4B-d0-vtok101-distr-base
Text Generation • 4B • Updated • 190
catt-bench/Qwen3.5-4B-d0-vtok101-lora-seeds
Updated
catt-bench/mt5-small-ftrace-retrain
Updated • 1
catt-bench/Qwen3.5-4B-route-ab-l8-lora-scale
Updated
catt-bench/Qwen3.5-4B-d0-vtok101-base
Text Generation • 4B • Updated • 318
datasets 6
catt-bench/route-attribution-baselines
Updated • 6.18k
catt-bench/vtok101-distr-attribution-baselines
Updated • 9.87k • 4
catt-bench/vtok101-attribution-baselines
Updated • 4.82k
catt-bench/ftrace-answer-only-query
Viewer • Updated • 200 • 4.88k
catt-bench/ftrace-trex-results
Updated • 66
catt-bench/catt-demo-data
Updated • 202