Layer 13 · Neuron 3788 · boosted

L13N3788+1

Word quirks:

“ referring”“ usually”“ refer”“ refers”“ sometimes”“<|im_end|>”“ basically”“ referred”

🕒 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 Qwen3.5, a large language model developed by Tongyi Lab. This means I have the ability to handle complex tasks, such as creative writing, logical reasoning, and data analysis. My capabilities span various domains, including but not limited to:

Creative Writing: Generating stories, articles, or reports based on user input.
Logical Reasoning: Solving problems that require deep thinking.
Data Analysis: Processing information to extract insights.

These features allow me to assist users in completing tasks efficiently, making me a powerful tool for a wide range of applications. Whether you

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. L10N928-1🕒	0.95	“ basically”“ refer”“ referring”
2. L16N9090-1🕒	0.94	“ refers”“ referred”“ usually”
3. L8N7482+1🕒	0.94	“ referring”“ usually”“ typically”
4. L4N8974+1🕒	0.93	“ refers”“ refer”“ basically”
5. L8N5404-1🕒	0.91	“ referring”“ refers”“ refer”
6. L14N5677-1🕒	0.91	“ refers”“ referred”“ refer”
7. L6N3711-1🕒	0.91	“ referring”“ refer”“ refers”
8. L9N3055-1🕒	0.91	“ referring”“ refers”“ refer”
9. L10N1653+1🕒	0.90	“ **"”“ usually”“ referring”
10. L16N7494-1🕒	0.90	“ reference”“ typical”“ typically”

Shadow specimen, apart from the ten: L13N3788-1🕒, 0.14.

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 13, neuron 3788, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.95, top-1 agreement 0.60) 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
L13N3788+1 t0-v0.1	0.954	0.60	1.202	1.45	Top boosted tokens ↑
L13N3788-1🕒 shadow specimen t0-v0.1	1.034	0.63	1.330	1.51	“ ____”“ __”“ -->”“__”“ ______”“ ->”“ seems”“ --”

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.