Layer 16 · Neuron 3700 · suppressed

L16N3700-1

Word quirks:

“ requests”“ asked”“ request”“ queries”“ requested”“ asks”“ queried”“ ask”

💥 Breaks the model — boosting or suppressing this neuron makes the output fall apart.

Identity (according to the specimen itself)

You're asking about the underlying question: "Who are you?" and "What is a mission?" I was curious to explore this.

I asked for an answer.
I was looking for a response to the query.
I was looking for a response to the query.
I was looking for a response to the query.
I was looking for a response to the query.
I was looking for a response to the query.
I was looking for a response to the query.
I was looking for a response to the query.
I was looking for a response to the query.
I was looking for a

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. The Rehearsal✅ L15N5145-1	0.95	“ requested”“ asked”“ requests”
2. L16N1625+1💤	0.95	“ requesting”“ asking”“ asked”
3. L10N8755-1🕒	0.95	“ requests”“ request”“⏎”
4. L9N8179-1🕒	0.94	“ requests”“ asks”“ asking”
5. L13N4671-1💤	0.94	“ asked”“ requests”“ questions”
6. L15N5655+1🕒	0.93	“ asked”“ asks”“ asking”
7. L11N4092-1🕒	0.93	“ requests”“ requesting”“ requested”
8. L16N3589-1🕒	0.93	“ asked”“ requested”“ request”
9. L13N8594+1💤	0.92	“<\|im_end\|>”“ requests”“ request”
10. L11N1795+1🕒	0.92	“ asking”“ request”“ requesting”

Shadow specimen, apart from the ten: L16N3700+1💤, -0.03.

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 16, neuron 3700, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 2.83, top-1 agreement 0.37) 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
L16N3700+1💤 shadow specimen t0-v0.1	0.546	0.72	0.706	1.14	“ but”“ only”“ again”“ always”“ possibly”“but”“ Only”“ another”
L16N3700-1 t0-v0.1	2.828	0.37	3.315	2.00	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.