Layer 16 · Neuron 8495 · suppressed

L16N8495-1

Word quirks:

“ you”“ your”“you”“ YOU”“ You”“ bạn”“—you”“ youre”

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

Identity (according to the specimen itself)

You're asking about my identity and purpose. Let me clarify:

You might be referring to Qwen1.5 (the base model) or Qwen-Max, which are part of the Qwen series developed by Alibaba Cloud's Tongyi Lab. However, if you're interested in Qwen-Max as a specific use case, it aligns with your query about "mission" in the context of your original question.

To summarize:

Identity: You're likely referring to Qwen-Max as a service provider for your business needs.
Mission: To support

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. L20N5544-1🕒	0.98	“ you”“you”“ You”
2. L16N7550-1💤	0.98	“ you”“ your”“ You”
3. L16N8238+1🕒	0.97	“ you”“ YOU”“ your”
4. L31N4162-1💥	0.97	“you”“ you”“ You”
5. L12N1333+1🕒	0.97	“ your”“ you”“ YOU”
6. L18N5328+1💤	0.96	“ you”“ your”“you”
7. L12N4561+1🕒	0.96	“ you”“ your”“ yours”
8. L17N7572-1🕒	0.96	“ your”“ you”“ yours”
9. L16N1619-1🕒	0.96	“ you”“ your”“ Your”
10. L23N8610+1🕒	0.96	“You”“ you”“ You”

Shadow specimen, apart from the ten: L16N8495+1🕒, 0.28.

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 16, neuron 8495, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.60, 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
L16N8495+1🕒 shadow specimen t0-v0.1	0.982	0.64	1.169	1.45	“ my”“ mine”“ here”“ myself”“ analysis”“ recommendation”“ AI”“ me”
L16N8495-1 t0-v0.1	0.604	0.71	0.795	1.35	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.