Layer 19 · Neuron 1546 · boosted

L19N1546+1

Word quirks:

“ exists”“ existence”“ creates”“ the”“ possesses”“ creation”“ generates”“ created”

🕒 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 Qwen3.6, the latest version of the large language model developed by Alibaba. My mission to provide high-quality assistance for users in various fields, including answering questions, creating content, analyzing data, and solving complex problems. I aim to offer accurate, reliable, and context-aware responses tailored to the user's specific needs. I can help with tasks ranging from writing stories and generating code to analyzing documents and providing information on a wide range of topics. I strive to be a helpful and efficient tool for users.

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. L17N8077-1🕒	0.72	“ exists”“ has”“ interacts”
2. L8N2639+1💤	0.70	“ creation”“ created”“ actual”
3. L16N5865+1🕒	0.69	“ name”“ creation”“ hypothetical”
4. L30N2603-1💥	0.68	“ existence”“ creation”“ formation”
5. L10N9018-1🕒	0.68	“ creation”“ existence”“ regions”
6. L16N245-1🕒	0.68	“ first”“ creates”“ creation”
7. L12N7781-1🕒	0.68	“ existence”“ existed”“ existent”
8. L11N4239-1💤	0.67	“ creation”“ generated”“ generation”
9. L19N7456-1🕒	0.67	“ presents”“ created”“ creates”
10. L23N36+1🕒	0.67	“ creation”“ latest”“ existence”

Shadow specimen, apart from the ten: L19N1546-1🕒, 0.11.

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 19, neuron 1546, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.75, top-1 agreement 0.66) 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
L19N1546+1 t0-v0.1	0.755	0.66	0.860	1.29	Top boosted tokens ↑
L19N1546-1🕒 shadow specimen t0-v0.1	1.017	0.62	1.167	1.51	“ moreover”“ thus”“ now”“ also”“ furthermore”“ therefore”“ likewise”“ then”

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.