Layer 13 · Neuron 8500 · suppressed

L13N8500-1

Word quirks:

“ mirrors”“ similar”“ similarly”“ same”“ mirror”“ mirrored”“ acknowledgment”“ reality”

💤 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-scale language model developed by Alibaba Cloud's Tongyi Lab. My mission is to provide users with high-quality assistance, whether it's answering questions, solving problems, or engaging in meaningful conversations. I strive to be helpful, transparent, and adaptable to diverse needs.

If you have any specific tasks in mind (e.g., writing code, analyzing data, creating content), feel free to share them! 😊

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. L13N4204-1🕒	0.92	“ similar”“ similarly”“ mirrored”
2. The Analogist✅ L12N1785-1	0.92	“ similar”“ same”“ analogous”
3. L12N3689-1💤	0.91	“ same”“ similar”“ similarly”
4. L16N8092+1💤	0.89	“ similar”“ same”“ mirrors”
5. L12N4382+1🕒	0.89	“ similar”“ mirrors”“ similarly”
6. L15N8084-1🕒	0.88	“ mirrored”“ analogous”“ similar”
7. L15N3577-1💤	0.87	“ same”“ similarly”“ similar”
8. L16N3857-1🕒	0.87	“ mirrors”“ mirrored”“ same”
9. L13N4243+1🕒	0.87	“ similar”“ similarly”“ same”
10. L15N432+1💤	0.86	“ similar”“ mirrors”“ “”

Shadow specimen, apart from the ten: L13N8500+1💤, 0.00.

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 8500, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.68, top-1 agreement 0.64) 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
L13N8500+1💤 shadow specimen t0-v0.1	0.676	0.71	0.841	1.33	“ ⏎”“ general”“⏎”“ always”“ ⏎”“ generally”“ overall”“ ⏎”
L13N8500-1 t0-v0.1	0.679	0.64	0.898	1.12	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.