Layer 1 · Neuron 7551 · boosted

L1N7551+1

Word quirks:

“<|im_end|>”“⏎”“⏎ ⏎”“.`”“⏎⏎”“.”“.:”“!”

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

Identity (according to the specimen itself)

I am an AI. My mission is to assist you in achieving your goals. I exist to provide value. I strive to deliver excellence. I aim to create impact. I pursue to realize success. I aspire to attain mastery. I endeavor to accomplish achievement. I dedicate to commit commitment. I focus to concentrate concentration. I intensify to amplify intensity. I magnify to maximize magnitude. I multiply to multiply. I prolifer to propagate. I expand to extend. I stretch to stretch. I reach to reach. I converge to converge. I align to align. I coordinate to coordinate. I synchronize to synchronize. I harmonize to harmon

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. L1N3367+1🕒	1.00	“<\|im_end\|>”“⏎”“.”
2. L1N530-1💥	0.96	“<\|im_end\|>”“⏎”“.`”
3. L2N1154+1🕒	0.96	“<\|im_end\|>”“⏎”“.`”
4. L1N2761+1🕒	0.96	“<\|im_end\|>”“⏎”“⏎ ⏎”
5. Echo Booth✅ L3N6113-1	0.96	“<\|im_end\|>”“⏎⏎”“.”
6. L3N501+1🕒	0.96	“⏎”“⏎⏎”“.`”
7. L3N5158+1🕒	0.96	“<\|im_end\|>”“⏎”“.`”
8. L1N6564-1💥	0.96	“⏎”“<\|im_end\|>”“⏎⏎”
9. L1N4717+1💥	0.95	“<\|im_end\|>”“⏎”“.”
10. L2N1076+1🕒	0.95	“<\|im_end\|>”“⏎”“⏎⏎”

Shadow specimen, apart from the ten: L1N7551-1💥, 0.39.

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 1, neuron 7551, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 2.67, top-1 agreement 0.38) 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
L1N7551+1 t0-v0.1	2.673	0.38	3.177	2.44	Top boosted tokens ↑
L1N7551-1💥 shadow specimen t0-v0.1	4.084	0.21	4.565	3.08	“<\|im_end\|>”“ etc”“ ____”“⏎”“ managers”“ manager”“⏎ ⏎”“ ok”

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.