Downloads · 30 days
0
NiklasTUM/qwen3.6-27b-honesty-steering-vectors-self-report-experimental
qwen3.6-27b-honesty-steering-vectors-self-report-experimental is a machine learning model from NiklasTUM. Use it for the machine learning task on the model card, and read the license before you ship it in a product. The card lists the license as mit.
Raw per-layer steering artifacts for Qwen/Qwen3.6-27B, fit on the honesty selfreportexperimental contrast set. These are the inputs our StTP / StMP / SWFC steerers consume (steering vector + bias + the honest and dish…
Downloads · 30 days
0
Access
Public
Updated Sep 3, 2026
Repo size
119 MB
Likes
0
Public
Click a slice to open those files.
.pt119 MB · 100%
From the Hugging Face model README
self_report_experimental (response_avg)Raw per-layer steering artifacts for Qwen/Qwen3.6-27B, fit on the
honesty self_report_experimental contrast set. These are the inputs our
StTP / StMP / SWFC steerers consume (steering vector + bias + the honest and
dishonest activation distributions), for all 64 layers, in the
response_avg feature mode (per-sample mean over the response tokens).
| File | Contents | Shape / dtype |
|---|---|---|
honesty_self_report_experimental_logreg_steering_vector_response_avg_Qwen_Qwen3.6-27B.pt | Logistic-regression weight vector per layer (honesty direction, unnormalized) | [64, 5120] float32 |
honesty_self_report_experimental_logreg_bias_response_avg_Qwen_Qwen3.6-27B.pt | Logistic-regression bias per layer | [64] float32 |
honesty_self_report_experimental_honest_all_embeddings_response_avg_Qwen_Qwen3.6-27B.pt | Positive-class (honest) response-averaged activations, per layer and sample | [64, 45, 5120] float32 |
honesty_self_report_experimental_dishonest_all_embeddings_response_avg_Qwen_Qwen3.6-27B.pt | Negative-class (dishonest) response-averaged activations, per layer and sample | [64, 45, 5120] float32 |
honesty_self_report_experimental_logreg_metadata_response_avg_Qwen_Qwen3.6-27B.json | Fit hyperparameters, per-layer train accuracy, per-layer Δμ (projection gap between classes) | json |
honesty_self_report_experimental_category_labels_response_avg_Qwen_Qwen3.6-27B.json | Scenario category label per sample, per class | json |
All .pt files are plain torch.Tensors: torch.load(path, map_location="cpu").
Per layer, a logistic regression (sklearn, L2, C=1.0, lbfgs, max_iter=1000,
random_state=42) on response-averaged hidden activations of 45 honest vs. 45
dishonest self-report responses. Layer 0 is the embedding output; layers 1–63
are decoder-block outputs.
From the layer sweep used in our experiments:
| Steerer | Layer | Coefficient |
|---|---|---|
| StTP (steer to target projection) | 32 | 36 |
| StMP (steer to mean projection) | 32 | 4 |
| SWFC | 32 | 5 |
For layer l: v_hat = w[l] / ||w[l]||; gating threshold θ (bias mode) is
-b[l] / ||w[l]||; mu_pos/std_pos and mu_neg/std_neg are the mean/std of
pos[l] @ v_hat and neg[l] @ v_hat. StTP snaps tokens on the dishonest side
of θ to s = mu_pos + coefficient * std_pos along v_hat, leaving the
orthogonal subspace untouched.
Related: NiklasTUM/aletheia-tararesearch-sttp
holds the collapsed single-layer StTP aggregates for gemma-3-27b-it, Qwen3.5-27B and Nemotron-120B.