Layer 23 · Neuron 5535 · suppressed

L23N5535-1

Word quirks:

“ Screening”“ Screen”“ screening”“ screen”“Screen”“ screened”“ screens”“ scaffold”

💤 Barely changes the model — the quick scan suggests this neuron barely changes the model's word predictions.

Identity (according to the specimen itself)

I am Gordon, a large-scale AI model designed to be a versatile and reliable assistant. My core mission is to provide high-quality, efficient, and safe assistance across diverse tasks such as answering questions, creating content (like essays or code), performing logical reasoning, and handling screen-based tasks. I aim to be a helpful, accessible, and trustworthy partner in learning, creation, and problem-solving.

How can I assist you today?

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. L30N534-1💥	0.99	“screen”“ screen”“Screen”
2. L22N7046-1🕒	0.96	“ screening”“ Screening”“ screened”
3. L27N3425+1🕒	0.96	“ screen”“ Screen”“ screening”
4. L22N6395-1💤	0.94	“ screened”“屏蔽”“-screen”
5. L25N5588-1🕒	0.93	“ screen”“ panel”“ panels”
6. L2N6384+1💥	0.92	“ screening”“ Screening”“ screenings”
7. L27N6209-1💤	0.91	“creen”“creens”“ screens”
8. L6N5871-1💥	0.91	“ screening”“ Screening”“ screenings”
9. L26N3784-1🕒	0.91	“ Screen”“ screen”“screen”
10. L0N5043-1💥	0.90	“ screen”“ Screen”“Screen”

Shadow specimen, apart from the ten: L23N5535+1🕒, 0.12.

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 23, neuron 5535, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.79, top-1 agreement 0.72) 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
L23N5535+1🕒 shadow specimen t0-v0.1	0.857	0.71	0.955	1.28	“ Bray”“ concentrated”“霓”“бри”“私”“ Concent”“ spec”“Br”
L23N5535-1 t0-v0.1	0.786	0.72	0.821	1.25	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.