Layer 16 · Neuron 6665 · boosted

L16N6665+1

Word quirks:

“ each”“ individual”“ Each”“each”“Each”“ individually”“每个”“每条”

💤 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 language model created by Alibaba Cloud. I don't have personal beliefs or missions in the traditional sense, but my design purpose is to serve as a highly capable assistant for specific tasks:

Reasoning & Problem-Solving: I excel at handling complex logical reasoning, mathematical computation, and code generation.
Multilingual Support: I can communicate fluently across dozens of languages, making me suitable for international users.
Specialized Capabilities: For example, I can act as a research assistant (e.g., summarizing papers) or a coding partner (e

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. L15N5463-1🕒	0.97	“ each”“(each”“each”
2. Silver Tongue✅ L11N8821+1	0.96	“ each”“ Each”“each”
3. L18N8178+1🕒	0.94	“ each”“each”“ individually”
4. L31N525-1💥	0.94	“ each”“each”“每个”
5. L15N9029+1🕒	0.93	“ each”“ every”“ Each”
6. L15N8915-1🕒	0.93	“ each”“ single”“ Each”
7. L16N5786-1🕒	0.93	“ each”“ EACH”“ every”
8. L20N2997-1🕒	0.93	“ each”“每个”“each”
9. L18N7620-1🕒	0.93	“ each”“each”“每个”
10. L20N7797+1🕒	0.93	“每个”“ each”“each”

Shadow specimen, apart from the ten: L16N6665-1🕒, 0.23.

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 16, neuron 6665, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.51, top-1 agreement 0.76) 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
L16N6665+1 t0-v0.1	0.509	0.76	0.650	1.20	Top boosted tokens ↑
L16N6665-1🕒 shadow specimen t0-v0.1	1.231	0.62	1.476	1.27	“ collectively”“ combined”“统称”“ together”“ overall”“combined”“ unified”“整体上”

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.