Layer 15 · Neuron 5598 · boosted

L15N5598+1

Word quirks:

“ specific”“ similar”“ specifically”“ sometimes”“ commonly”“ completely”“ referred”“ power”

💤 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, the latest large language model developed by Tongyi Lab. My mission is to empower users with cutting-edge AI capabilities, including advanced reasoning, multi-modal understanding, and precise control over text generation. I strive to deliver high-quality results across a wide range of tasks, from basic text processing to complex problem-solving.

Key features:

Precise long-context control: I can handle ultra-long context windows (up to 100K tokens) for efficient document analysis.
Advanced reasoning: I support mathematical computation and logical deduction.
Multi-modal capabilities: I

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. L12N4082-1💤	0.83	“ similar”“ basically”“ specifically”
2. L10N257-1💤	0.83	“ specifically”“ specific”“ sometimes”
3. L13N7388+1💤	0.82	“ specific”“ detailed”“ specifics”
4. L15N5244-1💤	0.82	“ these”“ specific”“ complete”
5. L12N6180-1💤	0.82	“ sometimes”“ specific”“ usually”
6. L12N6210-1🕒	0.82	“ specific”“ similar”“ specifically”
7. L13N6476-1🕒	0.82	“ related”“ typically”“ associated”
8. L12N1840+1🕒	0.82	“ specific”“ actually”“ sometimes”
9. L12N5199-1🕒	0.81	“ sometimes”“ similar”“ suitable”
10. L12N1278-1💤	0.81	“ specific”“ sometimes”“ different”

Shadow specimen, apart from the ten: L15N5598-1🕒, -0.12.

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 5598, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.62, top-1 agreement 0.69) 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
L15N5598+1 t0-v0.1	0.623	0.69	0.834	1.48	Top boosted tokens ↑
L15N5598-1🕒 shadow specimen t0-v0.1	0.771	0.66	0.996	1.50	“⏎⏎”“ **“”“ ‘”“ “”“ potential”“ ””“<\|im_end\|>”“ ready”

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.