Layer 8 · Neuron 433 · suppressed

L8N433-1

Word quirks:

“ sim”“ simulation”“ simulated”“ ⏎”“ ⏎⏎”“ simulate”“⏎”“ simul”

💤 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 artificial intelligence assistant designed to provide high-quality information and support. My mission is to serve as a versatile knowledge partner, offering assistance in areas such as language translation, content creation, and problem-solving across various domains. I strive to deliver accurate, relevant, and engaging responses that help users achieve their goals efficiently.

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. L6N3456-1🕒	0.95	“ simulation”“⏎”“ sim”
2. L18N5363+1💤	0.95	“模拟”“ simulated”“ simulate”
3. L7N6425-1💤	0.93	“ simulated”“ simulation”“ sim”
4. L7N1767+1🕒	0.93	“ simulated”“ simulation”“ `”
5. L9N4587+1🕒	0.93	“ simulation”“ simulated”“ sim”
6. L7N2497-1🕒	0.92	“<\|im_end\|>”“ simulated”“ simulation”
7. L12N1725+1💤	0.92	“ simulated”“ simulation”“ simulator”
8. L6N4091+1🕒	0.91	“ simulation”“ simulated”“ simulate”
9. L8N3910-1🕒	0.91	“ simulation”“ simulated”“ simulations”
10. L7N9082+1💤	0.91	“ simulated”“ simulation”“ sim”

Shadow specimen, apart from the ten: L8N433+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 8, neuron 433, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.88, top-1 agreement 0.62) 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
L8N433+1🕒 shadow specimen t0-v0.1	1.519	0.49	1.722	1.70	“ displaying”“ monitoring”“ integration”“ tracking”“ viewing”“ displayed”“ analysis”“ central”
L8N433-1 t0-v0.1	0.882	0.62	1.037	1.30	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.