Layer 4 · Neuron 6428 · suppressed

L4N6428-1

Word quirks:

“ ‘”“⏎”“ '”“ referenced”“ name”“ reference”“ --”“‘”

💤 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 an AI assistant designed to provide accurate, helpful, and harmless information. My mission is to assist users with their questions by offering reliable, factual, and unbiased responses. I do not have personal beliefs, opinions, or agendas of my own. I exist solely to serve the needs of those who seek my help.

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. L4N3854+1💤	0.93	“ ‘”“ “”“ name”
2. L7N7801-1💤	0.92	“ ‘”“‘”“ name”
3. L5N6270-1🕒	0.92	“ ‘”“ “”“ ——”
4. The Exegete✅ L4N3757-1	0.91	“ ‘”“ reference”“⏎”
5. L7N5262+1💤	0.91	“ ‘”“ '”“ referenced”
6. L7N3141-1💤	0.91	“ ‘”“ '”“ my”
7. L4N3942+1💤	0.91	“ ‘”“ '”“ my”
8. L5N7558-1🕒	0.91	“⏎”“ '”“ ‘”
9. L5N2952+1💤	0.91	“ ‘”“ '”“ reference”
10. L5N7175+1💤	0.90	“⏎”“ reference”“ referenced”

Shadow specimen, apart from the ten: L4N6428+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 4, neuron 6428, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.72, top-1 agreement 0.67) 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
L4N6428+1🕒 shadow specimen t0-v0.1	2.418	0.38	2.907	2.61	“<\|im_end\|>”“ ceramic”“ glass”“ laminate”“ metal”“ wifi”“ products”“ temperature”
L4N6428-1 t0-v0.1	0.718	0.67	0.882	1.49	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.