Layer 7 · Neuron 2769 · suppressed

L7N2769-1

Word quirks:

“⏎”“ defined”“ associated”“ referenced”“ "”“ response”“ free”“ reference”

💤 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, an advanced large language model designed to assist users with complex tasks. My mission is to provide high-quality, accurate, and efficient support across a wide range of domains, including but not limited to:

Data Analysis: Processing and interpreting large datasets to extract meaningful insights.
Problem Solving: Addressing intricate challenges that require precise solutions.
Innovation: Driving forward-thinking ideas and implementing them effectively.
User-Centric Design: Ensuring all outputs are tailored to meet user needs and expectations.

How can I assist you today? 😊

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. L3N7094-1🕒	0.77	“<\|im_end\|>”“⏎”“ reference”
2. L8N4982+1💤	0.77	“⏎”“ aligned”“ response”
3. L4N4993+1💤	0.76	“ referenced”“ reference”“ open”
4. L5N517-1🕒	0.76	“ reference”“ response”“ referenced”
5. L5N5482+1🕒	0.75	“ solution”“ referring”“ name”
6. L7N6563-1💤	0.75	“ "”“ reference”“ referenced”
7. L8N1020+1💤	0.75	“ reference”“ search”“ related”
8. L7N8367+1💤	0.75	“ referenced”“ reference”“ ⏎”
9. L9N6815-1💤	0.74	“ referenced”“ based”“ according”
10. L7N966-1💤	0.74	“ referenced”“ response”“ referring”

Shadow specimen, apart from the ten: L7N2769+1🕒, 0.20.

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 7, neuron 2769, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.68, top-1 agreement 0.67) 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
L7N2769+1🕒 shadow specimen t0-v0.1	1.762	0.46	2.059	2.01	“ reason”“ saying”“ thoughtful”“ ——”“ —”“ consideration”“ thinking”“ indeed”
L7N2769-1 t0-v0.1	0.679	0.67	0.841	1.32	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.