interp-engine 1.2.0__tar.gz → 1.2.2__tar.gz

{interp_engine-1.2.0 → interp_engine-1.2.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.5
 Name: interp-engine
-Version: 1.2.0
+Version: 1.2.2
 Summary: A fast, standardized interpretability engine that supports most modern models and architectures. Powers Neuronpedia.
 Project-URL: Homepage, https://github.com/decoderesearch/interp-engine
 Project-URL: Repository, https://github.com/decoderesearch/interp-engine
@@ -38,13 +38,13 @@ Description-Content-Type: text/markdown
   🔗 <a href="https://interp-engine.org"><strong>interp-engine.org</strong></a>
 </p>
 
-`interp-engine` is an interpretability engine that is fast (>40x tok/s vs HF eager), standardized (34 'points'/addresses across architectures), and easy to use and debug. It powers all of [Neuronpedia](https://neuronpedia.org)'s inference (circuit tracing, j-lens, steering, NLAs, etc) and is checked for accuracy against HF Transformers and other engines.
+`interp-engine` is an interpretability engine that is fast (>40x tok/s vs HF eager), standardized (34 'points'/addresses across architectures), and easy to use and debug. It powers all of [Neuronpedia](https://neuronpedia.org)'s inference and is checked for accuracy against HF Transformers and other engines.
 
 <p align="center">
-  <img src="https://neuronpedia.s3.amazonaws.com/site-assets/ie-perf.png" alt="Tokens per second while capturing and generating: eager against IE-vLLM and IE-vLLM-freeze on qwen3.8-27b and deepseek-v4-flash-0731, 8 requests in flight" width="620">
+  <img src="https://neuronpedia.s3.amazonaws.com/site-assets/ie-perf.png" alt="Tokens per second while capturing and generating: eager against IE-vLLM and IE-vLLM-freeze on qwen3.8-27b and deepseek-v4-flash-0731, 8 requests in flight" width="100%">
 </p>
 <p align="center">
-  <img src="https://neuronpedia.s3.amazonaws.com/site-assets/interp-engine-demo.gif" alt="interp-engine demo gif" width="480">
+  <img src="https://neuronpedia.s3.amazonaws.com/site-assets/interp-engine-demo.gif" alt="interp-engine demo gif" width="100%">
 </p>
 
 This repo contains:
@@ -129,11 +129,23 @@ We verify correctness in two main ways:
 
 ## Why use an Interpretability Engine, instead of just having my AI code whatever it needs on the fly?
 
-Of course, software is easy to make these days. But as smart as as AIs are these days, there are benefits to leveraging a pre-written engine:
+Software is easy to make in 2026, here's why an engine makes sense:
 
 1. **Speed**: Get performance without sacrificing correctness.
-1. **Standardization + Verification**: Eliminate ambiguity when referring to points, plus a full test suite included.
-1. **Faster Dev / Fewer Tokens Used**: You could spend ten million tokens and have your AI write, test, and make production-ready an interpretability engine. Or you could just `pip install interp-engine[vllm]`.
+2. **Standardization + Verification**: Eliminate ambiguity when referring to points, plus a full test suite included.
+3. **Faster Dev / Fewer Tokens Used**: You could spend ten million tokens and have your AI write, test, and make production-ready an interpretability engine. Or you could just `pip install interp-engine[vllm]`.
+
+## Development / Contributing
+
+Activate the shared git hooks once per clone — they format staged Python, rebuild the generated files, and run CI's static checks before a push. Details in [CONTRIBUTING.md](CONTRIBUTING.md).
+
+```bash
+make hooks    # or: git config core.hooksPath .githooks
+```
+
+## Contact
+
+Bugs and feature requests belong in [issues](https://github.com/decoderesearch/interp-engine/issues). For anything else: [johnny@neuronpedia.org](mailto:johnny@neuronpedia.org).
 
 ## License
 

{interp_engine-1.2.0 → interp_engine-1.2.2}/README.md

@@ -7,13 +7,13 @@
   🔗 <a href="https://interp-engine.org"><strong>interp-engine.org</strong></a>
 </p>
 
-`interp-engine` is an interpretability engine that is fast (>40x tok/s vs HF eager), standardized (34 'points'/addresses across architectures), and easy to use and debug. It powers all of [Neuronpedia](https://neuronpedia.org)'s inference (circuit tracing, j-lens, steering, NLAs, etc) and is checked for accuracy against HF Transformers and other engines.
+`interp-engine` is an interpretability engine that is fast (>40x tok/s vs HF eager), standardized (34 'points'/addresses across architectures), and easy to use and debug. It powers all of [Neuronpedia](https://neuronpedia.org)'s inference and is checked for accuracy against HF Transformers and other engines.
 
 <p align="center">
-  <img src="https://neuronpedia.s3.amazonaws.com/site-assets/ie-perf.png" alt="Tokens per second while capturing and generating: eager against IE-vLLM and IE-vLLM-freeze on qwen3.8-27b and deepseek-v4-flash-0731, 8 requests in flight" width="620">
+  <img src="https://neuronpedia.s3.amazonaws.com/site-assets/ie-perf.png" alt="Tokens per second while capturing and generating: eager against IE-vLLM and IE-vLLM-freeze on qwen3.8-27b and deepseek-v4-flash-0731, 8 requests in flight" width="100%">
 </p>
 <p align="center">
-  <img src="https://neuronpedia.s3.amazonaws.com/site-assets/interp-engine-demo.gif" alt="interp-engine demo gif" width="480">
+  <img src="https://neuronpedia.s3.amazonaws.com/site-assets/interp-engine-demo.gif" alt="interp-engine demo gif" width="100%">
 </p>
 
 This repo contains:
@@ -98,11 +98,23 @@ We verify correctness in two main ways:
 
 ## Why use an Interpretability Engine, instead of just having my AI code whatever it needs on the fly?
 
-Of course, software is easy to make these days. But as smart as as AIs are these days, there are benefits to leveraging a pre-written engine:
+Software is easy to make in 2026, here's why an engine makes sense:
 
 1. **Speed**: Get performance without sacrificing correctness.
-1. **Standardization + Verification**: Eliminate ambiguity when referring to points, plus a full test suite included.
-1. **Faster Dev / Fewer Tokens Used**: You could spend ten million tokens and have your AI write, test, and make production-ready an interpretability engine. Or you could just `pip install interp-engine[vllm]`.
+2. **Standardization + Verification**: Eliminate ambiguity when referring to points, plus a full test suite included.
+3. **Faster Dev / Fewer Tokens Used**: You could spend ten million tokens and have your AI write, test, and make production-ready an interpretability engine. Or you could just `pip install interp-engine[vllm]`.
+
+## Development / Contributing
+
+Activate the shared git hooks once per clone — they format staged Python, rebuild the generated files, and run CI's static checks before a push. Details in [CONTRIBUTING.md](CONTRIBUTING.md).
+
+```bash
+make hooks    # or: git config core.hooksPath .githooks
+```
+
+## Contact
+
+Bugs and feature requests belong in [issues](https://github.com/decoderesearch/interp-engine/issues). For anything else: [johnny@neuronpedia.org](mailto:johnny@neuronpedia.org).
 
 ## License
 

{interp_engine-1.2.0 → interp_engine-1.2.2}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "interp-engine"
-version = "1.2.0"
+version = "1.2.2"
 description = "A fast, standardized interpretability engine that supports most modern models and architectures. Powers Neuronpedia."
 readme = "README.md"
 license = "Apache-2.0"