This release is a pre-release and may not be stable for production use.
Undercurrent
See the undercurrent before it surfaces.
Undercurrent is Wrynx's activation-probing platform for LLM inference. You declare which
activations to capture during generation, run probes on them (classifiers or running trajectory
scores), and let inline probes intervene mid-generation, for example by aborting. Observe-mode probes
run asynchronously and never block the response. It works with Hugging Face transformers and vLLM.
This release is a placeholder that reserves the package name. It contains no functionality. The first real release (0.1.0) is in preparation. Follow https://github.com/wrynx/undercurrent for updates.
Licensed under Apache-2.0.
1 maintainer
Metadata
1 maintainer
Release files for undercurrent 0.0.0a0
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | |
|---|---|---|---|
| undercurrent-0.0.0a0.tar.gz | 5.2 kB | Details |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Reset |
|---|---|---|---|---|
| undercurrent-0.0.0a0-py3-none-any.whl | Python 3 | none | any | Details |
Total release size: 11.2 kB
Release files / undercurrent-0.0.0a0.tar.gz
| Download URL | undercurrent-0.0.0a0.tar.gz |
|---|---|
| Size | 5.2 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
b661a0f146e3521974cfda576a8c3aa33a8fb401126293c30a8f0c0ff2fe2a03
|
|
BLAKE2b-256 checksum How to use checksums |
26e9fd0c13f9afb6153b4406bd58603af0f91879f669c9dd305a699db982d48f
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/7.0.0 CPython/3.14.4
|
Release files / undercurrent-0.0.0a0-py3-none-any.whl
| Download URL | undercurrent-0.0.0a0-py3-none-any.whl |
|---|---|
| Size | 6.0 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
f1fda756b26b8cf78ac7767fd2353b03b6d6e131b7fa1f1ae9563d16989e1db9
|
|
BLAKE2b-256 checksum How to use checksums |
6bab8a104efe9f2447716d7d56a03ab7d8b82678c4cee08a47b2d66fd3aba657
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/7.0.0 CPython/3.14.4
|
