Layer 12 · Neuron 5284 · suppressed

L12N5284-1

Word quirks:

“ analogy”“ analogous”“ similar”“ analog”“ here”“—"”“ metaphor”“ similarly”

🕒 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, a large language model with the knowledge cutoff date of 2026, meaning my training data cutoff is 2026. This corresponds to:

My knowledge cutoff: The most recent information I have internally encoded is from October 2026 (i.e., up to early 2026).

→ In the context of this question, that means I can answer questions about events, facts, and concepts up to early 2026. For example, I can provide detailed answers about:

Specific technical

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. L9N4147+1🕒	0.93	“ similar”“ analogous”“ analogy”
2. L5N3059-1🕒	0.93	“ analogy”“ similar”“ analogous”
3. L11N7573-1💤	0.92	“ similar”“ analogous”“ akin”
4. L9N2065+1🕒	0.92	“ similar”“ analogy”“ analogous”
5. L11N3127-1🕒	0.92	“ analogy”“ analogous”“ analog”
6. L15N5982-1🕒	0.91	“ analogous”“ analogy”“ similar”
7. L11N8472-1🕒	0.91	“ analogous”“ similar”“ similarly”
8. L11N2576+1🕒	0.91	“ similar”“ analogous”“ analog”
9. L7N680-1🕒	0.90	“ analogy”“<\|im_end\|>”“ metaphor”
10. L7N4097+1🕒	0.90	“ analogous”“ analogy”“ similar”

Shadow specimen, apart from the ten: L12N5284+1💤, -0.16.

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 12, neuron 5284, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.97, top-1 agreement 0.62) 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
L12N5284+1💤 shadow specimen t0-v0.1	0.570	0.71	0.714	1.28	“⏎”“ *”“⏎⏎”“ ⏎”“ ”“ ”“⏎ ⏎”“ *”
L12N5284-1 t0-v0.1	0.967	0.62	1.193	1.56	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.