Layer 12 · Neuron 3102 · suppressed

L12N3102-1

Word quirks:

“ now”“ presently”“ basically”“ currently”“ soon”“ here”“ "*”“ essentially”

🧪 In testing — Answered 0 of 8 prompts so far.

Identity (according to the specimen itself)

I am Qwen3.5, a large language model developed by Tongyi Lab. My core capabilities include advanced reasoning across multiple domains, enabling me to assist users in optimizing their work processes and enhancing their decision-making. Here's an overview of my key features:

Advanced Reasoning Capabilities: I can break down complex tasks into manageable steps or use AI tools to automate them.
Multimodal Input Handling: I support direct input of text, images, audio, video, etc., allowing for seamless integration with various AI tools.
High-Quality Output Generation: I ensure accuracy and

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. L13N7569-1🕒	0.90	“ now”“ currently”“ directly”
2. L11N7649-1🕒	0.90	“ now”“ presently”“ here”
3. L9N8054+1🕒	0.86	“ now”“ currently”“ here”
4. L7N2564+1🕒	0.85	“ now”“ currently”“ actually”
5. L12N9101+1💤	0.85	“ now”“ currently”“ here”
6. L11N6300-1🕒	0.84	“ nowadays”“ now”“ presently”
7. L10N846-1🕒	0.84	“ nowadays”“ now”“ presently”
8. L13N520+1💤	0.84	“ now”“ currently”“ later”
9. L14N8303+1💤	0.84	“ now”“ thus”“ basically”
10. L11N7235+1🕒	0.83	“ nowadays”“ currently”“ now”

Shadow specimen, apart from the ten: L12N3102+1💤, 0.04.

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 3102, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. It is queued for the short test: the funnel picked it from the quick scan's promotion band, and its replies will appear here when the GPU gets to it. 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
L12N3102+1💤 shadow specimen t0-v0.1	0.769	0.68	1.018	1.34	“ ⏎”“ **”“⏎”“ ⏎”“ *”“ --”“ ⏎”“ ⏎⏎”
L12N3102-1 t0-v0.1	1.150	0.59	1.485	1.55	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.