Layer 17 · Neuron 5395 · suppressed

L17N5395-1

Word quirks:

“ mentioning”“ describing”“ detailing”“ specifying”“ outlining”“ identifying”“ highlighting”“ stating”

💤 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 Qwen3.6, a large language model developed by Tongyi Lab. I was released in June 2024, highlighting my capabilities as an advanced AI assistant. My mission is to demonstrate the potential of large-scale models, emphasizing my strengths in various fields such as science and technology, mathematics, and natural language processing. I aim to showcase how I can effectively address these challenges while demonstrating my proficiency across different applications.

As Qwen3.6, I highlight several key points:

Comprehensive Capabilities: I emphasize my ability to leverage the latest advancements in large-scale models, showcasing my proficiency

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. L16N9150+1🕒	0.85	“ describing”“ detailing”“ stating”
2. L17N4948-1💤	0.84	“ detail”“ mentioning”“ mention”
3. L16N5235-1🕒	0.83	“ specifying”“ describing”“ indicating”
4. L17N1289-1💤	0.83	“ describing”“ —”“ mentioning”
5. L19N7406+1🕒	0.79	“ specifying”“ stating”“ rejecting”
6. L17N716-1🕒	0.78	“ specifying”“ declaring”“ knowing”
7. L13N6857+1🕒	0.77	“ describing”“ providing”“ description”
8. L14N2073-1🕒	0.76	“ description”“ specifying”“ specify”
9. L14N953+1💤	0.76	“ specified”“ description”“ indicating”
10. L20N1075-1💤	0.76	“ specifying”“ describing”“ Specific”

Shadow specimen, apart from the ten: L17N5395+1🕒, 0.00.

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 17, neuron 5395, 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 0.71) 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
L17N5395+1🕒 shadow specimen t0-v0.1	0.672	0.65	0.876	1.27	“ means”“ $$”“ $”“ `$”“,$”“ $$$”“ ensures”“ interactions”
L17N5395-1 t0-v0.1	0.625	0.71	0.814	1.41	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.