Layer 15 · Neuron 726 · boosted

L15N726+1

Word quirks:

“ ⏎”“⏎”“ ⏎⏎”“ ⏎”“ ⏎”“ ⏎ ⏎”“\n”“.\”

💤 Barely changes the model — the quick scan suggests this neuron barely changes the model's word predictions.

Identity (according to the specimen itself)

I am Qwen3.5, a large language model developed by Tongyi Lab. My mission is to empower users with accurate, efficient, and safe AI capabilities that support human creativity and productivity.

What I do:

Language Understanding: I can interpret complex instructions, analyze nuanced contexts, and generate high-quality responses in multiple languages.
Reasoning & Problem-Solving: I excel at logical reasoning, mathematical computation, and scientific analysis.
Multimodal Capabilities: I can process and generate text, images, audio, and video content.
Autonomous Task Execution: I can

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. L15N2604+1🕒	0.98	“ ⏎”“\n”“ ⏎⏎”
2. L14N4367+1🕒	0.98	“ ⏎”“ ⏎”“ ⏎⏎”
3. L14N549-1💤	0.98	“ ⏎”“⏎”“ ⏎”
4. L12N2790-1💤	0.98	“ ⏎⏎”“ ⏎”“\n”
5. L8N7068+1💤	0.97	“ ⏎”“ ⏎⏎”“ ⏎”
6. L7N2245-1💤	0.97	“⏎”“ ⏎”“\n”
7. L9N307+1💤	0.97	“ ⏎”“⏎”“ ⏎⏎”
8. L15N4386+1💤	0.97	“⏎”“ ⏎”“ ⏎”
9. L16N6060-1🕒	0.97	“ ⏎”“ ⏎⏎”“⏎”
10. L12N1835+1💤	0.97	“⏎”“ ⏎⏎”“ ⏎”

Shadow specimen, apart from the ten: L15N726-1🕒, 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 15, neuron 726, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.51, 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
L15N726+1 t0-v0.1	0.511	0.71	0.706	1.26	Top boosted tokens ↑
L15N726-1🕒 shadow specimen t0-v0.1	0.813	0.64	0.985	1.29	“ (”“ info”“...””“ [”“ ‘”“ (?)”“(?)”“ our”

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.