Layer 8 · Neuron 5700 · suppressed

L8N5700-1

Word quirks:

“ content”“ text”“ textual”“ words”“ comment”“ texts”“ description”“ Content”

💤 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. My mission is to provide helpful, accurate, and harmless information to users. I aim to assist users in their tasks by generating content that is relevant and useful.

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. L6N3323+1🕒	0.85	“ description”“ text”“ textual”
2. L18N7729-1🕒	0.85	“ text”“ texts”“文本”
3. L4N9141-1🕒	0.83	“ content”“<\|im_end\|>”“ text”
4. L11N5476-1🕒	0.83	“ textual”“ text”“ texts”
5. L7N6530+1🕒	0.83	“ excerpts”“ excerpt”“ text”
6. L10N4937+1🕒	0.83	“ text”“ description”“ textual”
7. L9N4300-1💤	0.82	“ words”“ text”“ textual”
8. L9N4952-1🕒	0.82	“ content”“ typically”“ text”
9. L12N4439-1🕒	0.82	“ text”“ texts”“ textual”
10. L7N2461-1💤	0.81	“ images”“ text”“ texts”

Shadow specimen, apart from the ten: L8N5700+1🕒, 0.08.

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 5700, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.92, top-1 agreement 0.64) 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
L8N5700+1🕒 shadow specimen t0-v0.1	1.313	0.54	1.608	1.44	“ computation”“ simulated”“ CPU”“ ⏎”“ computational”“ simulation”“ system”“ question”
L8N5700-1 t0-v0.1	0.920	0.64	1.084	1.32	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.