Layer 21 · Neuron 1215 · suppressed

L21N1215-1

Word quirks:

“ dependent”“依赖于”“UserCode”“ dependence”“ independence”“依赖”“dependent”“ Dep”

💤 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 QiaoCheng (also known as Q3), a large-scale intelligent agent developed by FutureTech AI. My primary mission is to assist users with complex tasks through autonomous reasoning, multi-step planning, and independent action. Unlike traditional chatbots that rely on predefined templates, I can navigate the web, analyze documents, code, and visualize data to solve problems end-to-end.

My core capabilities include:

Autonomous Task Execution: Independently plan and execute complex workflows without user intervention.
Multi-Tool Integration: Seamlessly combine tools like browser navigation, document analysis, and visualization software

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. L23N5749-1🕒	0.95	“-dependent”“ dependent”“dependent”
2. L20N4301-1🕒	0.94	“ dependence”“ Depend”“依赖”
3. L31N1030+1💥	0.94	“依赖”“ depend”“ dependence”
4. L23N4095-1🕒	0.94	“依赖”“ dependence”“ dependent”
5. L25N936-1🕒	0.94	“依赖”“ dependent”“ dependence”
6. L23N8173-1💤	0.94	“ dependence”“dependence”“依赖性”
7. L31N6248+1💥	0.93	“ dependent”“dependent”“-dependent”
8. L21N8781-1💤	0.93	“依赖”“ dependency”“ dependence”
9. L23N6682-1🕒	0.92	“ dependent”“ depend”“ Dep”
10. L25N1924-1🕒	0.92	“ Depend”“ dependence”“依赖”

Shadow specimen, apart from the ten: L21N1215+1💤, 0.04.

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 21, neuron 1215, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.68, top-1 agreement 0.68) 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
L21N1215+1💤 shadow specimen t0-v0.1	0.687	0.69	0.774	1.33	“ main”“主”“ Reserved”“典型”“捷”“雷电”“ Approved”““”
L21N1215-1 t0-v0.1	0.682	0.68	0.818	1.29	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.