Layer 9 · Neuron 6987 · suppressed

L9N6987-1

Word quirks:

“ !!”“ !”“!!”“!"”“!\”“⏎”“ !!!”“!”

💤 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 Cloud. My mission is to provide users with high-quality, accurate, and helpful information while adhering to ethical guidelines and safety standards.

Key Features:

Advanced Knowledge: I possess extensive knowledge across various domains, including technology, science, culture, and more.
Ethical AI: My responses are designed to be safe, reliable, and beneficial, ensuring that all content aligns with positive values.
User-Centric Approach: I aim to assist users in making informed decisions based on the latest trends and insights.

###

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. L5N6129+1🕒	0.98	“!”“?!”“ !”
2. L10N8381-1🕒	0.98	“!”“!!”“!"”
3. L3N6174-1🕒	0.98	“!”“⏎”“!!”
4. L4N7970-1🕒	0.97	“!”“ !”“!"”
5. L5N940+1🕒	0.97	“!”“⏎”“<\|im_end\|>”
6. L6N8190+1💤	0.97	“!”“<\|im_end\|>”“⏎”
7. L4N8832+1🕒	0.97	“!”“!!”“ !!”
8. L5N3428+1🕒	0.97	“!”“?!”“!"”
9. L4N7499+1🕒	0.97	“ !”“!”“ !!”
10. L9N839-1💤	0.97	“!”“!"”“!\”

Shadow specimen, apart from the ten: L9N6987+1🕒, -0.07.

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 9, neuron 6987, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.86, top-1 agreement 0.61) 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
L9N6987+1🕒 shadow specimen t0-v0.1	2.268	0.46	2.619	1.51	“ always”“ any”“ forever”“ ALWAYS”“ permanently”“ if”“ anytime”“ never”
L9N6987-1 t0-v0.1	0.858	0.61	1.071	1.48	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.