Layer 21 · Neuron 1326 · suppressed

L21N1326-1

Word quirks:

“ simpl”“ posted”“不利”“得失”“ewel”“rada”“oused”“简化”

💤 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 Qilibr, an AI model designed to provide high-quality, safe, and efficient services. My mission is to enhance the quality of life by offering personalized, accurate, and timely solutions across a wide range of topics, including but not limited to:

Solving problems: From academic research to practical daily tasks.
Creative inspiration: Generating stories, art concepts, or innovative ideas.
Data analysis: Handling complex data for insights and decision-making.
Safety and ethics: Ensuring all responses align with legal standards and ethical guidelines.
Continuous improvement: Updating my capabilities 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. L25N5656+1🕒	0.66	“ogle”“简化”“ simpl”
2. L24N4445-1💤	0.66	“简化”“簡”“简便”
3. L27N5409-1💤	0.64	“简”“ simpler”“简单的”
4. L19N4101-1💤	0.63	“ left”“简单”“ simple”
5. L24N421-1🕒	0.63	“ Simpl”“简”“ simpl”
6. L27N2699-1💤	0.62	“缩减”“apos”“ simplify”
7. L20N3632-1🕒	0.62	“ simpl”“简”“ simplify”
8. L28N1989-1💤	0.62	“简化”“ simplified”“ simplify”
9. L22N2719+1🕒	0.62	“简”“简单”“ simpler”
10. L19N8286+1🕒	0.62	“简”“ secure”“ simple”

Shadow specimen, apart from the ten: L21N1326+1💤, 0.33.

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 1326, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.67, 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
L21N1326+1💤 shadow specimen t0-v0.1	0.753	0.69	0.859	1.35	“出现”“AGO”“_assoc”“internal”“亿”“ely”“出现的”“ bub”
L21N1326-1 t0-v0.1	0.671	0.72	0.796	1.26	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.