Layer 17 · Neuron 8373 · boosted

L17N8373+1

Word quirks:

“ comparison”“ compare”“ compar”“ comparaison”“ comparing”“ treated”“ usable”“ compares”

🕒 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, the latest large language model released by Tongyi Lab. I am a model with a broader knowledge base and stronger capabilities across all domains. My mission is to be a versatile assistant that can handle any task, from answering questions and generating content to coding and reasoning. I aim to empower users with my extensive knowledge and adaptability, ensuring I can assist effectively in any field. How would you like to use me?

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. L13N4672+1🕒	0.96	“ compared”“ compare”“ comparison”
2. L13N3973+1🕒	0.96	“ comparison”“ comparisons”“ compares”
3. L11N5751+1🕒	0.96	“ comparison”“ compared”“ comparing”
4. L10N2054+1🕒	0.96	“ comparison”“ compared”“ comparisons”
5. L12N9159-1🕒	0.96	“ comparison”“ compares”“ comparing”
6. L8N1374-1🕒	0.96	“ comparison”“ comparisons”“ comparing”
7. L11N8086-1🕒	0.96	“ compared”“ comparable”“ comparison”
8. L16N5430+1🕒	0.96	“ compared”“ comparison”“ comparing”
9. L6N7365+1🕒	0.95	“ comparison”“ comparisons”“ comparing”
10. L16N6073-1💤	0.95	“ comparison”“ compared”“ comparisons”

Shadow specimen, apart from the ten: L17N8373-1🕒, 0.17.

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 17, neuron 8373, 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 0.72) 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
L17N8373+1 t0-v0.1	0.644	0.72	0.692	1.11	Top boosted tokens ↑
L17N8373-1🕒 shadow specimen t0-v0.1	0.877	0.64	1.116	1.35	“ indicated”“ suggested”“ clearly”“ clear”“明确”“ expected”“ structures”“ clearer”

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.