Layer 16 · Neuron 373 · suppressed

L16N373-1

Word quirks:

“ again”“ similar”“ same”“ DT”“ info”“ update”“ another”“ definition”

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

Identity (according to the specimen itself)

Let's break this down:

Identity: I'm Qwen3.5, a large language model designed to handle complex tasks like reasoning, coding, or data analysis.
Mission: To provide accurate, reliable information while adhering to strict ethical guidelines. This includes maintaining high-quality performance across diverse scenarios, which aligns with my core capabilities as of 2026.

This is a simplified version of the information provided in the knowledge cutoff date. Let me know if you'd like more details! 😊

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. L13N3660+1💤	0.77	“ again”“ same”“ Again”
2. L14N3154+1🕒	0.72	“ same”“ info”“ SAME”
3. L17N701-1🕒	0.72	“ resolution”“ same”“ similar”
4. L14N4756+1🕒	0.71	“ same”“ again”“ another”
5. L13N3270-1🕒	0.71	“ again”“ same”“ similar”
6. L12N8499-1🕒	0.70	“ similar”“ similarly”“ related”
7. L16N100+1🕒	0.70	“ another”“ similarly”“ similar”
8. L16N3164+1🕒	0.69	“ similar”“ Again”“ Another”
9. L17N352+1🕒	0.69	“ info”“ same”“ similar”
10. L12N7082+1🕒	0.69	“ similar”“ similarly”“ same”

Shadow specimen, apart from the ten: L16N373+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 16, neuron 373, 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
L16N373+1🕒 shadow specimen t0-v0.1	0.656	0.65	0.945	1.26	“<\|im_end\|>”“⏎”“⏎ ⏎”“ ⏎”“ explaining”“ explanation”“ logic”“␍⏎”
L16N373-1 t0-v0.1	1.008	0.61	1.177	1.73	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.