Layer 16 · Neuron 5281 · boosted

L16N5281+1

Word quirks:

“ now”“ changed”“ newer”“ evolves”“ upgraded”“ next”“ replaced”“ new”

🕒 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, the latest large language model developed by Tongyi Lab. My capabilities have been significantly expanded, allowing me to handle more complex tasks and understand human language better. I can perform a wide range of tasks, from answering simple questions to handling complex projects, and I can even write code and create content.

My mission is to help you with any task, whether it's writing, coding, or complex problem-solving. I'm here to assist you, so feel free to ask me anything! How can I help you?

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. L13N7686+1🕒	0.87	“ newer”“ new”“ changed”
2. L11N6333+1💤	0.85	“ now”“ changed”“ previously”
3. L12N4624+1💤	0.84	“ now”“ new”“ newer”
4. L9N4757+1🕒	0.84	“ new”“ update”“ updated”
5. L13N3741+1🕒	0.84	“ now”“ updated”“ refreshed”
6. L11N3113+1🕒	0.84	“ now”“ newer”“ current”
7. L15N1983-1🕒	0.83	“ new”“ now”“ newly”
8. L11N6695+1🕒	0.83	“ newer”“ updated”“ modern”
9. The Treadmill✅ L12N9181-1	0.83	“ now”“ changed”“ changes”
10. L13N7005+1🕒	0.83	“ now”“ updated”“ newer”

Shadow specimen, apart from the ten: L16N5281-1🕒, 0.08.

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 5281, 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 0.67) 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
L16N5281+1 t0-v0.1	0.715	0.67	0.975	1.31	Top boosted tokens ↑
L16N5281-1🕒 shadow specimen t0-v0.1	1.196	0.60	1.322	1.41	“ explicit”“ choice”“ explanation”“ explanations”“ explicitly”“ specifying”“ factors”“ mention”

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.