Layer 11 · Neuron 2442 · suppressed

L11N2442-1

Word quirks:

“ here”“ precisely”“ this”“ particular”“ ⏎”“ certain”“"!”“!”

🕒 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 Qwen3.5, an advanced large language model developed by Alibaba Cloud. My mission is to assist users in solving complex problems by providing accurate information, generating creative content, or automating specific tasks. I aim to enhance user experience by integrating cutting-edge technology with practical applications. How can I help you today? Let's explore this together!

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. L14N7698-1💤	0.73	“ specific”“⏎”“ particular”
2. L10N1075-1💤	0.72	“ here”“ ⏎”“ detail”
3. L12N1504+1🕒	0.71	“ here”“ this”“ specifically”
4. L15N2987+1💤	0.70	“ specifically”“ here”“ particularly”
5. L12N8931-1🕒	0.69	“ specific”“ here”“ particular”
6. L13N7819-1💤	0.69	“ specific”“ particular”“ specifics”
7. L6N2003-1🕒	0.68	“ my”“ myself”“ ‘”
8. L12N3200-1💤	0.68	“ precisely”“ here”“ clarification”
9. L12N6758+1🕒	0.67	“ specific”“ particular”“ *”
10. L11N3786+1🕒	0.67	“ precise”“ precisely”“ exact”

Shadow specimen, apart from the ten: L11N2442+1🕒, 0.18.

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 11, neuron 2442, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.98, top-1 agreement 0.65) 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
L11N2442+1🕒 shadow specimen t0-v0.1	0.909	0.68	1.075	1.22	“ always”“ info”“ sometimes”“ ⏎”“ seems”“ description”“ ⏎ ⏎”“ knowing”
L11N2442-1 t0-v0.1	0.977	0.65	1.210	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.