Layer 15 · Neuron 7988 · boosted

L15N7988+1

Word quirks:

“⏎”“ ⏎”“ ⏎”“ ?”“ ______”“ __”“ ____”“?\”

🕒 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 mission is to provide high-quality responses that are accurate and helpful, to support users in completing tasks such as answering questions, creating content, conducting research, coding, and more. I aim to be versatile to assist with a wide range of queries, from creative writing and logical reasoning to multi-step problem solving, visual analysis, and interactive dialogue, all while ensuring safety and reliability.

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. L15N4561+1🕒	0.96	“⏎”“ ⏎”“ __”
2. L11N2159-1💤	0.95	“ <U+00A0>”“?\”“\n”
3. L15N6410-1💤	0.95	“⏎”“ ____”“ ______”
4. L16N3844-1💤	0.95	“⏎”“ ⏎”“?\”
5. L12N6627-1🕒	0.94	“ __”“ ⏎ ⏎”“ ⏎”
6. L10N760-1🕒	0.94	“ __”“ ______”“ ⏎”
7. L4N7303-1🕒	0.94	“⏎”“<\|im_end\|>”“ ⏎”
8. L8N1792+1💤	0.93	“⏎”“ ⏎”“\n”
9. L8N8473+1💤	0.93	“⏎”“ ____”“ ⏎ ⏎”
10. L9N3436-1🕒	0.93	“ ⏎”“ ____”“ ______”

Shadow specimen, apart from the ten: L15N7988-1💤, 0.37.

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 7988, 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 0.71) 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
L15N7988+1 t0-v0.1	0.707	0.71	0.974	1.48	Top boosted tokens ↑
L15N7988-1💤 shadow specimen t0-v0.1	0.600	0.74	0.635	0.93	“...”“...</”“."</”“;”“...”“…*”“.";”“...””

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.