Layer 8 · Neuron 6408 · suppressed

L8N6408-1

Word quirks:

“ reference”“ ‘”“ referenced”“ description”“ referring”“ “”“ references”“ specific”

💤 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, a next-generation large-scale AI model. My mission is to provide high-quality, accurate, and safe information to users while continuously advancing the capabilities of artificial intelligence in a responsible manner.

In terms of technical architecture, I have evolved beyond simple pattern matching. Instead of merely listing capabilities, I focus on delivering concrete, actionable insights that drive genuine progress. This approach ensures that my development remains aligned with human values and societal well-being, rather than just accumulating superficial features.

My core objective is to foster a healthy digital ecosystem where innovation is measured by its positive impact on humanity, rather than just raw

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. L8N5798+1💤	0.94	“ referenced”“ “”“ ‘”
2. L6N5238-1💤	0.93	“ referenced”“ reference”“ acknowledgment”
3. L5N7263+1🕒	0.92	“ referenced”“ reference”“ “”
4. L4N5883-1🕒	0.92	“ “”“ reference”“ referenced”
5. L7N6079-1💤	0.92	“ referenced”“ ‘”“ references”
6. L5N3893-1🕒	0.91	“ ‘”“ reference”“ referenced”
7. L5N3239-1🕒	0.91	“ reference”“ referenced”“⏎”
8. L4N2684-1🕒	0.91	“ referenced”“ ‘”“ referring”
9. L4N9150+1💤	0.90	“ reference”“ referenced”“ referring”
10. L16N1987+1💤	0.90	“ referenced”“ references”“ mentioned”

Shadow specimen, apart from the ten: L8N6408+1🕒, -0.09.

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 8, neuron 6408, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.94, top-1 agreement 0.59) 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
L8N6408+1🕒 shadow specimen t0-v0.1	1.451	0.54	1.612	1.50	“ Excel”“ spreadsheet”“ excel”“ tool”“ tools”“Excel”“ CSV”“ enabled”
L8N6408-1 t0-v0.1	0.937	0.59	1.157	1.49	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.