IAAR-Shanghai/Awesome-Attention-Heads
An awesome repository & A comprehensive survey on interpretability of LLM attention heads.
Provides a curated research platform with a peer-reviewed survey paper accepted by *Patterns* (Cell Press) that organizes attention head studies using a four-stage cognitive framework—Knowledge Recalling, In-Context Identification, Latent Reasoning, and Expression Preparation—to systematize mechanistic interpretability research. Includes structured paper taxonomy with experimental methodology classifications and causal analysis techniques (path patching, attribution heads, mediation analysis) for isolating functional circuits within transformer attention mechanisms across diverse LLM tasks.
400 stars. No commits in the last 6 months.
Stars
400
Forks
12
Language
TeX
License
—
Category
Last pushed
Mar 02, 2025
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/llm-tools/IAAR-Shanghai/Awesome-Attention-Heads"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
Lightning-AI/litgpt
20+ high-performance LLMs with recipes to pretrain, finetune and deploy at scale.
liangyuwang/Tiny-DeepSpeed
Tiny-DeepSpeed, a minimalistic re-implementation of the DeepSpeed library
microsoft/Text2Grad
🚀 Text2Grad: Converting natural language feedback into gradient signals for precise model...
catherinesyeh/attention-viz
Visualizing query-key interactions in language + vision transformers (VIS 2023)
FareedKhan-dev/Building-llama3-from-scratch
LLaMA 3 is one of the most promising open-source model after Mistral, we will recreate it's...