Layer 26 · Neuron 5114 · suppressed

L26N5114-1

Word quirks:

“ external”“ External”“ externa”“外部”“External”“external”“_external”“的外部”

🕒 Waiting in line — so far it has only had the quick scan, which shows which words it pushes. It was not selected in this round.

Identity (according to the specimen itself)

I am Q15, a large-scale external1000-level (1000 外) model developed by X1000 (X1000 外部). My primary mission is to provide high-quality, accurate, and helpful information across a wide range of topics—whether you need creative writing, logical reasoning, scientific analysis, or even heavy external knowledge retrieval.

I am designed to be your versatile assistant:

External Knowledge: I can access and synthesize information from the external web to answer questions about current events, global trends, and specialized topics.
**Reason

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. L27N2298-1💥	1.00	“ External”“External”“ external”
2. L27N7749-1🕒	1.00	“ External”“ external”“External”
3. L25N2932+1💥	0.99	“外部”“ external”“ External”
4. L22N5726-1🕒	0.99	“ external”“外部”“external”
5. L23N2096+1🕒	0.99	“ external”“外部”“ External”
6. L27N6598+1🕒	0.98	“ External”“外”“ external”
7. L26N3466-1💤	0.98	“ External”“External”“ extern”
8. L29N8233-1💥	0.98	“外部”“ external”“ External”
9. L26N6963-1🕒	0.97	“ external”“外部”“external”
10. L24N4403+1🕒	0.96	“ extern”“ external”“ External”

Shadow specimen, apart from the ten: L26N5114+1🕒, -0.01.

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 26, neuron 5114, 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 0.72) 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
L26N5114+1🕒 shadow specimen t0-v0.1	0.877	0.69	0.830	1.44	“ D”“ cl”“[D”“clare”“ коп”“-cl”“*D”“ cla”
L26N5114-1 t0-v0.1	0.722	0.72	0.765	1.19	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.