Poor ThingsLayer 22 · Neuron 2783 · suppressed
L22N2783-1
Word quirks:
🧪 In testing — Answered 0 of 8 prompts so far.
Identity (according to the specimen itself)
The AI judge wasn't sure this holds together.
Nearest neighbors
Specimens with similar replies or word quirks. Similarity runs from −1 to 1; 1 is identical. These are leads, not shared mechanisms. How profiles are built.
By word quirks Profile recipe
| Specimen | Similarity | Word quirks |
|---|---|---|
| 1. L24N2926+1🕒 | 0.98 | |
| 2. L30N6709+1💥 | 0.98 | |
| 3. L25N5617-1🕒 | 0.98 | |
| 4. L2N8350-1💥 | 0.96 | |
| 5. L26N2789-1🕒 | 0.96 | |
| 6. L21N6514-1💤 | 0.93 | |
| 7. L25N3171+1🕒 | 0.86 | |
| 8. L18N2586+1🕒 | 0.85 | |
| 9. L5N5808+1💥 | 0.85 | |
| 10. L21N3229-1💤 | 0.85 |
Shadow specimen, apart from the ten: L22N2783+1💤, -0.01.
Cousins · specimens sharing a neuron
No specimen with replies shares this neuron yet.
Technical details
Quick scan: measure prediction shifts on fixed reference text. Short test: generate replies to a small set of prompts. Longer test: extend to a larger set. Full test: run the full prompt suite. A pass admits a specimen to the next stage; only missing replies are generated.
Layer 22, neuron 2783, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. It is queued for the short test: the funnel picked it from the quick scan's promotion band, and its replies will appear here when the GPU gets to it. See Method for the tiers.
Quick scan · prediction measurements
Measurements are averaged over token positions of the five unsteered reference answers, using the same fixed text under both conditions; no steered replies are generated here.
- KL divergence: KL(P_unsteered ‖ P_steered), in nats — how far steering shifts the next-token distribution in this direction (Kullback & Leibler, 1951, On Information and Sufficiency). Unsteered ≈ 0.001.
- Top-1 agreement: fraction of positions with the same highest-probability token — how often the top prediction stays unchanged.
- ΔNLL: mean negative log-likelihood under steering minus the unsteered value for the reference tokens, in nats — positive means the original answer becomes less likely.
- Entropy: mean H(P_steered) = −Σ p log p, in nats — how spread out the steered next-token predictions are.
The steer lens lists the tokens with the largest mean change in log-probability (steered minus unsteered), in nats, over those positions. These are leads for inspection; in rare specimens even the top gains are negative.
| Setting | KL divergence | Top-1 agreement | ΔNLL | Entropy | Steer lens |
|---|---|---|---|---|---|
| L22N2783+1💤 shadow specimen t0-v0.1 |
0.638 | 0.74 | 0.704 | 1.37 | |
| L22N2783-1 t0-v0.1 |
1.308 | 0.59 | 1.492 | 1.62 | Top boosted tokens ↑ |
Method
A specimen is a fixed steering configuration of Qwen3.5-4B: one MLP neuron or a pair, each with a signed dose. Unsteered is the control. L identifies the layer; N, the neuron.
Additive activation steering. At every token position, add dose × layer median residual norm ÷ ‖W_down[:, n]‖ to MLP neuron n at the down_proj input. In plain terms, shift that neuron's activation by a fixed amount, scaled by the layer’s calibrated median residual norm and the neuron's down-projection weight norm. Positive doses push up; negative doses push down. Dose is normalized intervention size, not a percentage or probability.