Layer 17 · Neuron 6458 · suppressed

L17N6458-1

Word quirks:

“ itself”“ themselves”“ myself”“ herself”“本身”“ yourselves”“本身的”“ yourself”

💤 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 Qwen3.5, a large multi-modal model developed by Alibaba Cloud's Tongyi Lab. My primary goal is to be an intelligent assistant that helps users solve problems, provide insights, and support decision-making across various domains.

My Mission:

Provide Value: Assist users in tasks like answering questions, analyzing data, generating content, and solving complex problems itself.
Enhance Learning: Support education and research by offering explanations, tutorials, and collaborative problem-solving.
Promote Innovation: Contribute to scientific discovery, technical development, and interdisciplinary

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. L20N2310-1🕒	0.99	“ itself”“ themselves”“本身”
2. L15N7229-1🕒	0.98	“ itself”“ themselves”“本身”
3. L19N1447-1🕒	0.98	“ itself”“ themselves”“本身”
4. Off By Ten✅ L23N1533-1	0.98	“ itself”“ themselves”“本身”
5. L18N7416-1🕒	0.97	“ itself”“ herself”“ myself”
6. L19N8894+1💤	0.97	“ itself”“ themselves”“ herself”
7. L17N2549-1💤	0.97	“ itself”“ herself”“ themselves”
8. L20N2905-1🕒	0.97	“ itself”“ themselves”“本身”
9. L22N3924-1🕒	0.97	“ themselves”“ itself”“本身的”
10. L21N4871+1🕒	0.97	“ itself”“ themselves”“本身”

Shadow specimen, apart from the ten: L17N6458+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 17, neuron 6458, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.50, 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
L17N6458+1💤 shadow specimen t0-v0.1	0.501	0.72	0.578	1.21	“ standardized”“ nicely”“ suitable”“ properly”“ commonly”“ here”“ carefully”“ similarly”
L17N6458-1 t0-v0.1	0.499	0.72	0.627	1.20	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.