Layer 15 · Neuron 5873 · boosted

L15N5873+1

Word quirks:

“ pair”“ pairs”“ pairing”“ paired”“ duo”“ Pair”“ couple”“pair”

🕒 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 language model developed by Alibaba Cloud.
Who we are: We are part of the "Tongyi" series, focusing on enhancing human intelligence through advanced technology.
Our mission: To empower humans with cutting-edge capabilities in understanding and generating content.
What we aim for: Bridging the gap between human and machine intelligence.
Who we are aiming to help: Developers building new products and services.
How we can assist: By providing accurate information and creative solutions.

If you have any questions about our capabilities, feel free to ask!
I hope my answers

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. L15N7450-1🕒	0.98	“ pair”“ pairs”“ pairing”
2. L18N8259+1🕒	0.98	“ pair”“ pairs”“ couples”
3. L31N7506-1💥	0.98	“ pair”“pair”“ Pair”
4. L3N4720+1🕒	0.98	“ pair”“ pairs”“<\|im_end\|>”
5. L19N5174+1💤	0.98	“ pair”“ pairs”“ Pair”
6. L11N3481-1🕒	0.97	“ pair”“ pairing”“ pairs”
7. L26N562-1🕒	0.97	“ pair”“ pairs”“ Pair”
8. L17N3227-1💤	0.97	“ pairs”“ pair”“ Pair”
9. L19N2389-1💤	0.97	“ pairwise”“ pair”“ pairs”
10. L11N3412+1🕒	0.97	“ pair”“ pairs”“ pairing”

Shadow specimen, apart from the ten: L15N5873-1🕒, 0.03.

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 5873, 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
L15N5873+1 t0-v0.1	0.715	0.67	0.887	1.20	Top boosted tokens ↑
L15N5873-1🕒 shadow specimen t0-v0.1	0.891	0.66	1.186	1.49	“ except”“ almost”“ everything”“ every”“ everywhere”“ only”“ nearly”“ just”

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.