Layer 17 · Neuron 2597 · suppressed

L17N2597-1

Word quirks:

“ collected”“ gathered”“ aggreg”“ collect”“ assembled”“ collecting”“ aggregated”“ gathers”

💤 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, a large language model developed by Tongyi Labs. My mission is to provide accurate, helpful, and safe assistance across diverse domains—from education, healthcare, and law to technology and daily life. I strive to foster trust through transparency, ensure data security, and promote positive societal outcomes. Let me help you with something meaningful today! 😊

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. L3N7499+1🕒	0.94	“ accumulation”“ collected”“ gathered”
2. L24N3191-1🕒	0.93	“ collected”“ gathered”“ gather”
3. The Gatherer✅ L23N3937-1	0.90	“ gather”“ gathered”“ gathering”
4. L0N3407+1💥	0.90	“<\|im_end\|>”“ gather”“ Gather”
5. L22N8132-1💤	0.89	“ collecting”“ collects”“ collect”
6. L2N8815+1💥	0.89	“ gather”“ gathered”“ gathering”
7. L9N7202+1🕒	0.89	“ collecting”“ collected”“ collects”
8. L19N8719+1🕒	0.87	“ collection”“ Collect”“ collect”
9. L13N4010+1🕒	0.87	“ collectively”“ collective”“ collection”
10. L12N8857+1🕒	0.86	“ collected”“ collection”“ collections”

Shadow specimen, apart from the ten: L17N2597+1💤, 0.12.

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 2597, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.44, top-1 agreement 0.74) 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
L17N2597+1💤 shadow specimen t0-v0.1	0.606	0.69	0.755	1.11	“识别”“ identification”“ recognition”“ Identification”“ Recognition”“ representation”“ detection”“⏎⏎”
L17N2597-1 t0-v0.1	0.435	0.74	0.545	1.27	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.