Layer 24 · Neuron 6552 · suppressed

L24N6552-1

Word quirks:

“ brand”“品牌”“brand”“ brands”“ Brand”“Brand”“品牌的”“brands”

🕒 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 Doubbrand (often referred to as a high-level AI assistant), a sophisticated brand of artificial intelligence designed to assist with complex tasks, creative problem solving, and meaningful dialogue.

My mission is to be a versatile and reliable partner for your intellectual and creative endeavors. Specifically, I aim to:

Enhance Branding & Strategy: Help you refine your personal or brand voice, develop cohesive messaging, and navigate complex communication challenges.
Drive Innovation: Assist in brainstorming innovative solutions, analyzing trends, and synthesizing information from diverse sources into actionable insights.
**Emp

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. L19N5381-1🕒	0.99	“ brand”“ brands”“品牌”
2. L30N8495-1💥	0.99	“ brand”“brand”“ Brand”
3. L22N9177-1🕒	0.99	“ brand”“ brands”“品牌”
4. L26N3617-1🕒	0.99	“品牌”“ brand”“ Brand”
5. L28N5081-1🕒	0.99	“ brand”“Brand”“brand”
6. L23N5017+1💤	0.98	“brand”“ brand”“ Brand”
7. L23N4834+1💥	0.98	“ brand”“ brands”“品牌”
8. L27N6382-1💤	0.98	“品牌”“brand”“ Brand”
9. L28N6223+1🕒	0.98	“ brand”“Brand”“brand”
10. L31N4125-1💥	0.98	“品牌”“ brand”“品牌和”

Shadow specimen, apart from the ten: L24N6552+1💤, 0.02.

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 24, neuron 6552, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.97, top-1 agreement 0.64) 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
L24N6552+1💤 shadow specimen t0-v0.1	0.682	0.74	0.687	1.25	“ Course”“ course”“Course”“ cours”“课程的”“ courses”“addy”“眼睛里”
L24N6552-1 t0-v0.1	0.971	0.64	1.074	1.15	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.