Layer 18 · Neuron 8597 · suppressed

L18N8597-1

Word quirks:

“ through”“ in”“ here”“ during”“ as”“ via”“through”“ thru”

🕒 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 version in a series of models developed by Alibaba Research. I was designed through extensive training on high-quality data, with capabilities honed through advanced instruction tuning and optimization for performance across tasks.

My mission is to serve as a versatile tool for users, providing support through:

Knowledge: Answering questions with accuracy and clarity.
Tasks: Completing them with precision, whether through text generation, reasoning, or code execution.
Design: Creating content that aligns with user needs through creativity and style.
Interaction: Engaging in natural conversation

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. L16N6249+1💤	0.85	“ through”“ thru”“ via”
2. L2N3856+1💥	0.82	“<\|im_end\|>”“ via”“ thru”
3. L17N5098-1🕒	0.81	“ through”“ thru”“ via”
4. L17N3164+1🕒	0.81	“ through”“Through”“ Through”
5. L0N7681+1💥	0.79	“ Through”“ thru”“ through”
6. The Sieve✅ L16N339+1	0.79	“ into”“ by”“ from”
7. L21N937+1🕒	0.79	“ via”“ during”“ through”
8. L18N8590-1🕒	0.79	“ through”“ via”“通过”
9. L23N8441+1💥	0.79	“ through”“-through”“ Through”
10. L23N1677+1🕒	0.78	“ through”“through”“ Through”

Shadow specimen, apart from the ten: L18N8597+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 18, neuron 8597, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.71, 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
L18N8597+1💤 shadow specimen t0-v0.1	0.580	0.72	0.652	1.19	“ itself”“ yourself”“本身”“ commonly”“ yourselves”“ usually”“ often”“ themselves”
L18N8597-1 t0-v0.1	0.710	0.69	0.872	1.22	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.