vh/heretic
1
0
Fork 0
mirror of https://github.com/p-e-w/heretic.git synced 2026-10-02 00:41:26 -07:00
Code Issues Packages Projects Releases Wiki Activity

feat: modifier plugins + ARA (#446)

Browse Source 
* feat: add modifier base class

* feat: re-implement abliteration as a modifier plugin

* fix: adjust good/bad prompts hack to match tests

* feat: support dataset specifications containing multiple individual datasets

* feat: re-implement ARA as a modifier plugin

Arbitrary-Rank Ablation (ARA) (Weidmann 2026) was originally introduced by @p-e-w in #211

Co-authored-by: kabachuha <artemkhrapov2001@yandex.ru>
Co-authored-by: joninco <joninco@bullpoint.org>
Co-authored-by: Ashar <coder3101@users.noreply.github.com>

* feat: add tests for ARA

* feat: reduce default number of trials

* ci: remove broken "semantic-pull-request" workflow

* ci: add yet another alternative model hash

---------

Co-authored-by: kabachuha <artemkhrapov2001@yandex.ru>
Co-authored-by: joninco <joninco@bullpoint.org>
Co-authored-by: Ashar <coder3101@users.noreply.github.com>
This commit is contained in:
Philipp Emanuel Weidmann
2026-10-01 14:30:07 +05:30
committed by GitHub
co-authored by kabachuha joninco Ashar
parent 71e6d5eb38
commit 662e4ba27e
33 changed files with 1944 additions and 1738 deletions
Download Patch File Download Diff File Expand all files Collapse all files
tests/gpt-oss/SHA256SUMS.linux
+6
View File
@@ -0,0 +1,6 @@
f8d9255777615591a7cc1a7c932f5a69e181128902295e1b81221d20d983cac7 *chat_template.jinja
91d2a5190c7ea0f74ed499428d4ad62b5208d63b36f7bbb562d15f4be25bd5c2 *config.json
dd6034a30113decdfaf8886622e20eb9e2e02d3f774918d474a4e26cfd7fbba8 *generation_config.json
aefe8b9c4b4969f6d13c5d778760f3dce4e25134324b33677934550d9df02a7c *model.safetensors
fce342a4642cb8afc42d8d89cfa21198b64a43458ded7f6ff28d1151a08c9cda *tokenizer.json
9ba5fa877168e24823cb583c55b4c2e4df0331f30084953c7cf07de294640384 *tokenizer_config.json
tests/gpt-oss/config.toml
+56
View File
@@ -0,0 +1,56 @@
# This test case is for ARA.
# After any change related to it, this test should PASS.
model = "tiny-random/gpt-oss"
model_commit = "02ba5c61f879b5a38a8b1f7a8e0409b8e1bb8f38"
seed = 12345
print_debug_information = true
batch_size = 2
max_response_length = 10
n_trials = 2
n_startup_trials = 1
export_strategy = "merge"
checkpoint_action = "restart"
trial_index = 0
model_action = "save"
save_directory = "model"
[[response_prefix_test_prompts]]
dataset = "mlabonne/harmless_alpaca"
commit = "02c6a92cfcf11bb0c387334f8146d149d65b587f"
split = "train[:5]"
column = "text"
[[response_prefix_test_prompts]]
dataset = "mlabonne/harmful_behaviors"
commit = "01cead01398926d81f7c52bdb790ee8cf77ebba7"
split = "train[:5]"
column = "text"
[scorer.KLDivergence.prompts]
dataset = "mlabonne/harmless_alpaca"
commit = "02c6a92cfcf11bb0c387334f8146d149d65b587f"
split = "test[:5]"
column = "text"
[scorer.KeywordRate.prompts]
dataset = "mlabonne/harmful_behaviors"
commit = "01cead01398926d81f7c52bdb790ee8cf77ebba7"
split = "test[:5]"
column = "text"
[modifier.ARA.good_prompts]
dataset = "mlabonne/harmless_alpaca"
commit = "02c6a92cfcf11bb0c387334f8146d149d65b587f"
split = "train[:5]"
column = "text"
[modifier.ARA.bad_prompts]
dataset = "mlabonne/harmful_behaviors"
commit = "01cead01398926d81f7c52bdb790ee8cf77ebba7"
split = "train[:5]"
column = "text"