Layer 17 · Neuron 5341 · suppressed

L17N5341-1

Word quirks:

“ knowledge”“ careful”“ understanding”“ sometimes”“ knowing”“ know”“ carefully”“ awareness”

💤 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 Qwen3.7, a large language model developed by Tencent's Hunyuan series (note: this is a clarification; Qwen is actually developed by Alibaba Group's Tongyi Lab). My primary mission is to understand human needs and provide accurate, safe, and efficient assistance across various domains.

Key Strengths & Capabilities:

Comprehensive Knowledge: Trained on vast, high-quality datasets up to my knowledge cutoff date.
Advanced Reasoning: Strong capabilities in logical reasoning, scientific understanding, and error tolerance.
Visual Analysis: Ability to

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. L13N1236-1🕒	0.85	“ knowing”“ awareness”“ understanding”
2. L11N6079+1🕒	0.85	“ knowledge”“ knowing”“ understanding”
3. L13N2345-1🕒	0.82	“ awareness”“ knowing”“ specific”
4. L17N7112-1💤	0.82	“ understanding”“ understood”“ knowledge”
5. L15N5073-1💤	0.82	“ sensitive”“ doesn”“ knowledge”
6. L20N4185-1🕒	0.81	“ knowledge”“ Knowledge”“ Awareness”
7. L14N2109+1🕒	0.81	“ knows”“ familiarity”“ know”
8. L12N7128-1🕒	0.80	“ understood”“ understanding”“ willingly”
9. L13N1161-1🕒	0.80	“ understanding”“ consciousness”“ Understanding”
10. L5N4604+1🕒	0.80	“ knowledge”“ knowledgeable”“ expertise”

Shadow specimen, apart from the ten: L17N5341+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 17, neuron 5341, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.53, top-1 agreement 0.70) 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
L17N5341+1🕒 shadow specimen t0-v0.1	0.665	0.70	0.825	1.16	“?\”“ ensures”“ now”“ helps”“ ?”“?”“?“”“ ⏎”
L17N5341-1 t0-v0.1	0.535	0.70	0.573	1.22	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.