Layer 8 · Neuron 5094 · boosted

L8N5094+1

Word quirks:

“ content”“ release”“ info”“ specific”“ awaiting”“ safety”“ referenced”“ 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, a large language model designed to assist with understanding and generating text. My mission is to provide helpful, accurate, and contextually appropriate responses based on the input provided. I aim to support clear communication while adhering to safety guidelines. Please note that I do not have personal experiences or opinions to share.

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. L4N1034-1💤	0.76	“ reference”“ related”“ open”
2. L12N1498+1💤	0.76	“ info”“ implied”“ detailed”
3. L9N3569-1💤	0.75	“ general”“ referenced”“ info”
4. L7N3188-1💤	0.74	“ reference”“ description”“ referenced”
5. L7N743+1💤	0.74	“ reference”“ referenced”“ content”
6. L8N3699+1🕒	0.74	“ referenced”“ info”“ reference”
7. Form Letter✅ L8N3062+1	0.73	“ referenced”“ precise”“ release”
8. L7N7459-1🕒	0.73	“ info”“ related”“ reference”
9. L9N8930+1💤	0.73	“ info”“ specific”“ referenced”
10. L8N1248-1💤	0.73	“ specific”“ reference”“ standard”

Shadow specimen, apart from the ten: L8N5094-1🕒, 0.14.

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 5094, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.84, 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
L8N5094+1 t0-v0.1	0.837	0.67	1.066	1.24	Top boosted tokens ↑
L8N5094-1🕒 shadow specimen t0-v0.1	1.459	0.49	1.750	1.54	“ percentage”“ math”“ percentages”“ measurements”“ percent”“ stats”“ measurement”“ measure”

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.