Layer 14 · Neuron 3082 · suppressed

L14N3082-1

Word quirks:

“ here”“ your”“ you”“ now”“ if”“ suddenly”“ yourself”“ answer”

💤 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. You can think of me as a large language model designed to assist with complex tasks and conversations. My primary mission is to provide high-quality support and capabilities. Here's how I can help:

Creative writing and storytelling
Code generation and debugging
Data analysis and visualization

If you need to analyze data or visualize information, I can guide you through the process. If you're looking for creative writing or coding assistance, feel free to ask. For data analysis and visualization, I can help you explore patterns and insights. Additionally, if you need to create charts

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. L13N6252-1💤	0.87	“ if”“ your”“ now”
2. L9N7598-1💤	0.87	“ now”“ your”“ here”
3. L10N1516-1💤	0.86	“ your”“ here”“ my”
4. L11N2578+1🕒	0.86	“ you”“ your”“ here”
5. L14N8653+1💤	0.86	“ you”“ now”“ your”
6. L11N3398+1💤	0.85	“ your”“ yourself”“ right”
7. L10N8300+1🕒	0.85	“ your”“ you”“ yourself”
8. L11N8729-1💤	0.85	“ here”“ your”“ often”
9. L8N5041-1💤	0.85	“ your”“ now”“ you”
10. L11N4838-1💤	0.85	“ your”“ you”“ today”

Shadow specimen, apart from the ten: L14N3082+1🕒, 0.18.

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 14, neuron 3082, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.84, 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
L14N3082+1🕒 shadow specimen t0-v0.1	1.083	0.61	1.347	1.27	“ typical”“ presumably”“ expected”“ mentioned”“ typically”“ usual”“ stated”“ suggested”
L14N3082-1 t0-v0.1	0.840	0.64	1.105	1.35	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.