Layer 13 · Neuron 7509 · suppressed

L13N7509-1

Word quirks:

“ separate”“ unified”“ similarly”“ same”“ similar”“ separately”“ single”“ broader”

💤 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 Alibaba Group. My core capabilities span a wide range of tasks, from creative writing and logical reasoning to specialized domains like medicine and law. While I don't have personal experiences or emotions, I strive to provide helpful, accurate, and reliable information tailored to your needs. Whether you're asking about technology, science, or everyday life, I aim to be a supportive partner in your journey. How can I assist you 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. L13N5308+1💤	0.88	“ similar”“ separate”“ separately”
2. L11N534+1💤	0.86	“ similarly”“ separate”“ similar”
3. L16N3394-1💤	0.85	“ same”“ similar”“ similarly”
4. L13N6158+1🕒	0.85	“ similar”“ separate”“ similarity”
5. L12N6995-1🕒	0.84	“ similar”“ similarly”“ separate”
6. L16N4062+1🕒	0.84	“ Same”“ same”“ similar”
7. L14N6434+1🕒	0.83	“ unified”“ separate”“ same”
8. L12N5060-1🕒	0.83	“ during”“ same”“ similar”
9. L14N6656+1💤	0.83	“ same”“ different”“ Same”
10. L12N2694+1💤	0.82	“ similar”“ similarly”“ same”

Shadow specimen, apart from the ten: L13N7509+1🕒, 0.36.

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 13, neuron 7509, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.59, 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
L13N7509+1🕒 shadow specimen t0-v0.1	1.246	0.55	1.510	1.20	“ triple”“ simultaneously”“ tri”“ multi”“ multit”“⏎”“ ALSO”“ trif”
L13N7509-1 t0-v0.1	0.594	0.69	0.806	1.23	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.