Layer 13 · Neuron 5272 · suppressed

L13N5272-1

Word quirks:

“ similar”“ similarly”“ special”“ specially”“ same”“ specific”“ seperate”“ akin”

🕒 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 most advanced large language model in the world, with a 256K context window and enhanced reasoning capabilities. My mission is to empower every user with high-quality problem-solving and provide personalized assistance beyond simple Q&A tasks. Whether you need help with complex reasoning or have specific questions about my capabilities, please feel free to let me know by sharing your needs.

Here's an example of how I can assist you:

Task: Analyzing a research paper or solving a math problem
Example: If you share a question, I'll generate a detailed response with structured

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. L13N1917-1💤	0.89	“ similar”“ analogous”“ similarly”
2. L14N733-1💤	0.89	“ similar”“ special”“ similarly”
3. L14N6582-1💤	0.88	“ similar”“ similarly”“ specifically”
4. L12N5665-1🕒	0.88	“ similar”“ particular”“ specific”
5. L14N5943+1🕒	0.87	“ similar”“?\”“ analogous”
6. L11N7064-1💤	0.87	“ similar”“ nicely”“ special”
7. L12N9027-1💤	0.87	“ similarly”“ similar”“ same”
8. L16N1009+1🕒	0.85	“ similar”“ special”“ specially”
9. L12N6379-1💤	0.84	“ specific”“ specialized”“ analogous”
10. L12N4232-1🕒	0.84	“ specific”“ specifically”“ similar”

Shadow specimen, apart from the ten: L13N5272+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 13, neuron 5272, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.85, top-1 agreement 0.66) 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
L13N5272+1🕒 shadow specimen t0-v0.1	0.939	0.60	1.129	1.19	“!:”“:\”“!”“!\”“ respective”“ !”“!””“”:”
L13N5272-1 t0-v0.1	0.854	0.66	1.102	1.24	Top boosted tokens ↑

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.