Layer 12 · Neuron 5556 · boosted

L12N5556+1

Word quirks:

“ exact”“ specifics”“ direct”“ specific”“ specifically”“ precise”“ directly”“ basically”

💤 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, the latest large language model developed by Tongyi Lab. Here's a breakdown of my identity and mission:

---

Who Am I?

Identity: I am an advanced AI model designed to handle complex tasks with high precision.
Key Features:
Enhanced Language Foundation: Built on a robust pre-trained base, optimized for accuracy in text generation.
Advanced Reasoning: Capable of solving math problems or logical challenges through step-by-step analysis.
Full-Stack Development Support: Includes capabilities for end-to-end

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. L13N4441-1💤	0.93	“ directly”“ specific”“ direct”
2. L10N3267+1💤	0.92	“ specifically”“ specific”“ directly”
3. L12N3894-1💤	0.92	“ specific”“ exact”“ specifically”
4. L12N4477+1💤	0.91	“ specific”“ specifically”“ specifics”
5. L9N4164+1💤	0.91	“ specific”“ precise”“ Specific”
6. L12N5831+1🕒	0.91	“ exactly”“ precisely”“ exact”
7. L12N4047+1🕒	0.91	“ specific”“ exactly”“ directly”
8. L12N4974+1🕒	0.90	“ precisely”“ specifically”“ specific”
9. L11N4311+1🕒	0.90	“ exact”“ specific”“ specifics”
10. L9N8707+1🕒	0.90	“ specific”“ Specific”“ specifics”

Shadow specimen, apart from the ten: L12N5556-1🕒, -0.07.

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 5556, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.64, top-1 agreement 0.69) 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
L12N5556+1 t0-v0.1	0.641	0.69	0.736	1.25	Top boosted tokens ↑
L12N5556-1🕒 shadow specimen t0-v0.1	1.441	0.52	1.780	1.86	“ __”“ ____”“ ______”“ ___”“ [[”“ _____”“ _”“ --”

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.