Layer 12 · Neuron 6062 · boosted

L12N6062+1

Word quirks:

“ specific”“ specifically”“ generally”“ general”“ Specific”“specific”“ specifics”“ always”

🕒 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-scale AI language model developed by Alibaba Cloud. My capabilities span various domains, including natural language processing, code generation, and complex reasoning tasks. Here's a breakdown of my core functions:

Language Understanding: I can process text in over 100 languages, enabling clear communication and precise information retrieval.
Code Generation: I assist developers by writing code snippets, debugging, and creating applications, making it easier to build software and solve problems.
Advanced Reasoning: I excel at logical reasoning, mathematical calculations, and analyzing complex data, allowing me to tackle

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. L12N1926+1💤	0.95	“ specific”“ specifics”“ specifically”
2. L13N7052+1💤	0.95	“ certain”“ specifics”“ specific”
3. L11N8713-1💤	0.94	“ specific”“ specifically”“ Specific”
4. L10N7710-1🕒	0.94	“ specific”“ specifics”“ general”
5. L11N9079+1🕒	0.94	“ specific”“ particular”“ usually”
6. L11N8717-1🕒	0.93	“ specific”“ specialized”“ Specific”
7. L11N3161+1🕒	0.93	“ specific”“ unspecified”“ specifically”
8. L9N7335+1🕒	0.93	“ specific”“ specifics”“ Specific”
9. L11N6235+1🕒	0.92	“ generally”“ general”“ specific”
10. L8N5448+1💤	0.92	“ specific”“ specifically”“⏎”

Shadow specimen, apart from the ten: L12N6062-1🕒, 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 12, neuron 6062, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.80, 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
L12N6062+1 t0-v0.1	0.801	0.64	1.001	1.22	Top boosted tokens ↑
L12N6062-1🕒 shadow specimen t0-v0.1	0.944	0.65	1.193	1.40	“ your”“ similar”“ similarly”“ you”“ here”“ focus”“ Your”“ related”

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.