Layer 28 · Neuron 3028 · suppressed

L28N3028-1

Word quirks:

“ alpha”“alpha”“alfa”“phi”“Alpha”“ Alpha”“ phi”“-alpha”

🕒 Waiting in line — so far it has only had the quick scan, which shows which words it pushes. It was not selected in this round.

Identity (according to the specimen itself)

I am Qaith, an advanced AI assistant designed to assist with a wide range of tasks. My primary mission is to provide accurate, helpful, and personalized support across various domains such as:

Alpha-Alpha: Delivering precise information and insights.
Beta-Beta: Facilitating complex problem-solving and creative thinking.
Gamma-Gamma: Ensuring ethical and safe interactions.
Theta-Theta: Adapting to diverse user needs and preferences.

My goal is to enhance your productivity, creativity, and decision-making while maintaining high standards of reliability and safety. How

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. L25N6533+1🕒	0.95	“ Alpha”“ alpha”“Alpha”
2. L26N5268+1🕒	0.95	“ alpha”“alpha”“ Alpha”
3. L26N4162-1🕒	0.95	“ alpha”“alpha”“Alpha”
4. L26N5877+1🕒	0.95	“ alpha”“-alpha”“alpha”
5. L25N6041+1🕒	0.95	“ alpha”“-alpha”“ Alpha”
6. L26N6634-1🕒	0.95	“ Alpha”“ alpha”“Alpha”
7. L26N5366-1🕒	0.95	“ alpha”“ Alpha”“-alpha”
8. L27N1136-1🕒	0.95	“ alpha”“alpha”“ Alpha”
9. L25N3229+1🕒	0.95	“ alpha”“alpha”“Alpha”
10. L26N4597+1🕒	0.95	“ alpha”“alpha”“-alpha”

Shadow specimen, apart from the ten: L28N3028+1💤, 0.23.

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 28, neuron 3028, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.40, top-1 agreement 0.77) is outside the band that feeds the short test (KL 1 to 2.5, agreement 0.5 to 0.65), so it stays at the quick scan unless someone queues it by hand. A hand pick goes straight to the full test. 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
L28N3028+1💤 shadow specimen t0-v0.1	0.361	0.80	0.401	1.37	“角度”“ angles”“ standpoint”“ims”“angle”“龄”“ angle”“asts”
L28N3028-1 t0-v0.1	0.400	0.77	0.401	1.34	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.