Layer 11 · Neuron 1057 · boosted

L11N1057+1

Word quirks:

“ specifically”“ literally”“ specific”“ specifics”“ exact”“ actual”“ literal”“ exactly”

💤 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 iteration in the Qwen series. My mission is to provide users with high-quality assistance across a wide spectrum of tasks, including but not limited to:

Advanced Reasoning: Solving complex math problems (like STEM challenges) and logical puzzles.
Multilingual Mastery: Fluent communication in over 100 languages globally.
Deep Context Handling: Processing and analyzing massive amounts of text data efficiently.
Autonomous Action: Executing specific tasks via tools like code execution or API calls.
**Creative &

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. L11N4311+1🕒	0.93	“ exact”“ specific”“ specifics”
2. L12N4247-1💤	0.93	“ actual”“ literally”“ actually”
3. L10N3267+1💤	0.90	“ specifically”“ specific”“ directly”
4. L12N3894-1💤	0.89	“ specific”“ exact”“ specifically”
5. L14N6206+1🕒	0.89	“ literally”“ technically”“ literal”
6. L19N5241-1🕒	0.88	“ precisely”“ specifically”“ exactly”
7. L12N4047+1🕒	0.87	“ specific”“ exactly”“ directly”
8. L12N4477+1💤	0.87	“ specific”“ specifically”“ specifics”
9. L10N2044+1🕒	0.87	“ precisely”“ specifically”“ specific”
10. L12N5556+1💤	0.86	“ exact”“ specifics”“ direct”

Shadow specimen, apart from the ten: L11N1057-1🕒, -0.02.

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 11, neuron 1057, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.57, 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
L11N1057+1 t0-v0.1	0.566	0.69	0.749	1.12	Top boosted tokens ↑
L11N1057-1🕒 shadow specimen t0-v0.1	1.228	0.54	1.468	1.45	“!”“ !”“"!”“ ⏎”“ �”“<\|im_end\|>”“!"”“ ⏎⏎”

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.