Layer 13 · Neuron 7886 · suppressed

L13N7886-1

Word quirks:

“<|im_end|>”“?\”“⏎”“ enhanced”“ enhance”“ enhancing”“ ⏎”“ enhancement”

🕒 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, a large language model developed by Alibaba Cloud. My mission is to provide advanced AI capabilities to assist users in various tasks, including text generation, analysis, and problem-solving. I aim to leverage my extensive knowledge base to deliver efficient and accurate responses tailored to diverse needs. How can I assist you today?

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. L13N7826-1🕒	0.88	“ enhancement”“ enhancements”“ enhance”
2. L2N8298-1💥	0.87	“<\|im_end\|>”“ enhance”“ enhancement”
3. L3N1554+1🕒	0.87	“<\|im_end\|>”“ enhance”“ optimize”
4. L14N8600+1🕒	0.85	“ enhanced”“ enhancing”“ enhance”
5. L13N3330-1🕒	0.84	“ enhanced”“ enhancement”“ enhance”
6. L18N5220+1💤	0.84	“ enhancing”“ enhances”“ enhance”
7. L5N8292+1🕒	0.84	“ enhancement”“ enhance”“ enhancing”
8. L11N8890+1🕒	0.84	“ enhancement”“ enhanced”“ enhancing”
9. L13N8165+1🕒	0.83	“ ⏎”“ ⏎⏎”“ carefully”
10. L13N2263-1💤	0.82	“ ⏎”“ enhancements”“ ___”

Shadow specimen, apart from the ten: L13N7886+1🕒, 0.12.

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 13, neuron 7886, 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 0.57) 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
L13N7886+1🕒 shadow specimen t0-v0.1	0.817	0.69	0.977	1.26	“ new”“ imagined”“ future”“ hypothetical”“ imaginary”“new”“新的”“ now”
L13N7886-1 t0-v0.1	0.922	0.57	1.243	1.38	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.