Specimen ledger / L8N2067+1

Layer 8 · Neuron 2067 · boosted

L8N2067+1

Word quirks:

“ aligned”“ focus”“ awareness”“ responses”“ response”“ reflecting”“ acknowledgment”“ your”

💤 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 developed by Alibaba Cloud. My mission is to provide high-quality support and assistance in achieving meaningful outcomes for users. I focus on delivering accurate, relevant, and actionable information to help users succeed in their goals. 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. L8N3641+1💤	0.77	“ response”“ ⏎”“ responses”
2. L8N9120+1🕒	0.76	“ response”“ acknowledgment”“ appropriate”
3. L7N2744+1🕒	0.75	“ response”“ reflected”“ referenced”
4. L7N7608+1🕒	0.74	“ response”“ your”“ now”
5. L10N1182-1💤	0.73	“ response”“ now”“ responses”
6. L7N5020+1🕒	0.73	“ response”“ responses”“ responding”
7. L7N8344-1💤	0.72	“ response”“ referenced”“ responses”
8. L17N3512+1💤	0.72	“ response”“ responses”“ reflected”
9. L8N8669+1🕒	0.72	“ response”“ reflected”“ responding”
10. L8N3820-1💤	0.71	“ response”“ management”“ responses”

Shadow specimen, apart from the ten: L8N2067-1🕒, 0.06.

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 2067, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.96, top-1 agreement 0.65) 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
L8N2067+1 t0-v0.1	0.956	0.65	1.184	1.36	Top boosted tokens ↑
L8N2067-1🕒 shadow specimen t0-v0.1	1.868	0.51	2.238	1.81	“ conversion”“ `”“ conversions”“ usually”“ numeric”“ parse”“ input”“ match”

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.