Layer 16 · Neuron 8415 · boosted

L16N8415+1

Word quirks:

“ included”“ embedded”“ mini”“ contained”“ incorporated”“ embed”“ inclusion”“ amongst”

🕒 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 Qwen, a large-scale language model. My mission is to provide high-quality assistance in various domains, including but not limited to:

Task Execution: Completing complex tasks such as writing stories, creating games, and mathematical computations.
Multilingual Support: Serving over 100 languages and regions.
AI Alignment: Ensuring all content adheres to safety guidelines while delivering helpful responses.
Technology Integration: Incorporating advanced capabilities like full-stack code generation, multi-agent collaboration, and efficient planning within my knowledge cutoff of 2026.

I aim to balance

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. L15N7776+1🕒	0.88	“ embed”“ embedded”“ inclusion”
2. L12N8966-1🕒	0.87	“ included”“ includes”“ include”
3. L14N7390-1💤	0.87	“ included”“ inclusion”“ incorporated”
4. L17N3687+1🕒	0.85	“ included”“ inclusion”“ incorporated”
5. L19N6819+1🕒	0.85	“ included”“ embedded”“included”
6. L31N2058+1💥	0.85	“ inclusion”“ included”“included”
7. L25N3807-1🕒	0.84	“ included”“included”“ Included”
8. L19N3334+1💤	0.84	“ incorpor”“ incorporated”“embedded”
9. L15N7218+1🕒	0.83	“ included”“included”“ incorporated”
10. L19N6758+1🕒	0.83	“ inclusion”“ included”“ integral”

Shadow specimen, apart from the ten: L16N8415-1🕒, 0.25.

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 8415, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.65, 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
L16N8415+1 t0-v0.1	0.647	0.71	0.820	1.35	Top boosted tokens ↑
L16N8415-1🕒 shadow specimen t0-v0.1	0.744	0.71	0.897	1.09	“ separate”“ surrounding”“ nearby”“ Separate”“ separately”“ afterward”“ seperate”“ afterwards”

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.