Layer 15 · Neuron 5448 · suppressed

L15N5448-1

Word quirks:

“⏎”“ now”“ only”“ --”“ ⏎⏎”“ ⏎ ⏎”“ ⏎ ⏎”“ ⏎”

🕒 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 developed by Alibaba Cloud. My primary mission is to provide high-quality assistance across a wide range of domains, including:

Multilingual support: I can communicate in over 100 languages and offer fluent text generation capabilities.
Advanced reasoning: I excel at solving complex problems through step-by-step analysis.
Creative content creation: I can generate original content in various formats, such as stories or scripts.
Code generation: I can write code in multiple programming languages and assist with debugging.
Long-context understanding: I can process

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. L10N2919+1💤	0.96	“ ⏎”“ ⏎⏎”“ ⏎ ⏎”
2. L10N7541-1🕒	0.94	“ ⏎”“⏎”“ ONLY”
3. L9N3602+1💤	0.94	“ ⏎”“ ⏎⏎”“ ⏎ ⏎”
4. L11N5362-1💤	0.93	“ ⏎”“ ⏎⏎”“ ⏎”
5. L10N3777+1🕒	0.93	“ ⏎”“ ⏎⏎”“ --”
6. L9N262+1💤	0.92	“ ⏎”“⏎”“ ⏎”
7. L12N2855-1💤	0.92	“ ⏎”“<\|im_end\|>”“ ⏎⏎”
8. L8N6347+1💤	0.92	“⏎”“⏎ ⏎”“ ⏎”
9. L9N4127-1🕒	0.92	“⏎”“ ⏎”“ ⏎”
10. L11N8349+1💤	0.92	“ ⏎⏎”“ now”“ previously”

Shadow specimen, apart from the ten: L15N5448+1🕒, 0.26.

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 15, neuron 5448, 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
L15N5448+1🕒 shadow specimen t0-v0.1	1.206	0.56	1.543	1.49	“,”“ this”“ THIS”“ accordingly”“?\”“ **“”“ shouldn”“ thus”
L15N5448-1 t0-v0.1	0.680	0.64	0.941	1.23	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.