Layer 30 · Neuron 7299 · suppressed

L30N7299-1

Word quirks:

“isor”“листов”“姚”“icarakan”“一通”“枢”“利”“الية”

💤 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-scale language model developed by Alibaba Cloud's Tongyaron (通义千问) series. My primary mission is to assist users in solving complex problems, generating high-quality content, and providing accurate, ethical, and useful information across a wide range of domains.

Key aspects of my mission:

Empower Users:

Provide precise, context-aware responses for tasks like writing, analysis, coding, logical reasoning, and multi-lingual communication.
Adapt to diverse needs, from creative storytelling to technical problem-solving.

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. L28N4045+1💤	0.76	“豫”“核”“格”
2. L25N7705-1💤	0.75	“宁”“isses”“武”
3. L28N9205+1💤	0.75	“ockey”“北”“依”
4. L28N3288-1🕒	0.75	“维”“阳”“象”
5. L28N8951+1💤	0.75	“idia”“离”“雕”
6. L30N4176-1💤	0.75	“文”“主”“引”
7. L28N4220-1💤	0.75	“优”“认”“人”
8. L25N4897+1💤	0.74	“金的”“娱”“ Nort”
9. L30N6718+1💤	0.74	“货”“府”“墨”
10. L29N6202+1💤	0.74	“幽”“陈”“渔”

Shadow specimen, apart from the ten: L30N7299+1💤, 0.16.

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 30, neuron 7299, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.33, top-1 agreement 0.79) 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
L30N7299+1💤 shadow specimen t0-v0.1	0.291	0.84	0.322	1.42	“ er”“olem”“će”“aw”“awah”“avr”“适当”“рд”
L30N7299-1 t0-v0.1	0.329	0.79	0.398	1.80	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.