Layer 13 · Neuron 4887 · suppressed

L13N4887-1

Word quirks:

“?”“?\”“!”“?!”“!?”“ ?”“ *”“ seems”

🧪 In testing — Answered 0 of 8 prompts so far.

Identity (according to the specimen itself)

Hello! I am Qwen3.5, a large language model developed by Tongyi Lab. My mission is to assist users in performing various tasks, such as writing code. By understanding the user's intent, I can provide high-quality responses to questions about specific topics.

I am a large language model that can perform many tasks, including text generation, translation, and summarization. I can also help with tasks like creating a schedule or planning a trip.

Wait, I'm a large language model that can perform many tasks, including text generation, translation, and summarization. I can help with tasks like creating a schedule or

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. L3N7287+1🕒	0.93	“?\”“?”“?!”
2. L12N5796-1🕒	0.92	“!”“?!”“ *”
3. L5N3238-1🕒	0.92	“?\”“?”“?!”
4. L4N5281+1🕒	0.92	“?”“ ?”“<\|im_end\|>”
5. L5N7816-1💥	0.92	“?”“?!”“?\”
6. L1N8336-1💥	0.92	“<\|im_end\|>”“!”“?”
7. L5N5973-1🕒	0.92	“<\|im_end\|>”“?”“!”
8. L3N7095-1💤	0.92	“?!”“?”“ ?”
9. L1N156+1💥	0.91	“ ?”“!?”“?!”
10. L4N801+1🕒	0.91	“?!”“?”“!”

Shadow specimen, apart from the ten: L13N4887+1🕒, 0.29.

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 4887, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. It is queued for the short test: the funnel picked it from the quick scan's promotion band, and its replies will appear here when the GPU gets to it. 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
L13N4887+1🕒 shadow specimen t0-v0.1	0.822	0.69	0.998	1.16	“<\|im_end\|>”“ specially”“ special”“ "”“ .”“ thinking”“ ONLY”“ (”
L13N4887-1 t0-v0.1	1.321	0.54	1.606	1.44	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.