Specimen ledger / L15N2073+1

Layer 15 · Neuron 2073 · boosted

L15N2073+1

Word quirks:

“ “”“ ‘”“ ⏎”“⏎”“ same”“ ””“ similarly”“ ’”

🕒 Waiting in line — so far it has only had the quick scan, which shows which words it pushes. It was not selected in this round.

Identity (according to the specimen itself)

I am Qwen3.5, the latest iteration in Alibaba Cloud's large language model family. I am designed to be a versatile AI assistant tailored for diverse scenarios.

Who am I?
I am an advanced AI model developed by Alibaba Cloud, built on extensive training data.

What is my mission?
My mission is to serve as a general-purpose AI assistant that adapts to user needs across different domains. Whether you need help with complex tasks, creative writing, coding, or analysis, I aim to provide accurate, safe, and helpful support.

Key capabilities:

**Multi-modal

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. L12N4865+1💤	0.94	“ ‘”“ ⏎”“ ⏎”
2. L12N3844-1🕒	0.93	“ “”“ ⏎”“ ⏎”
3. L14N3084+1🕒	0.93	“ “”“ similarly”“ similar”
4. L11N6093+1💤	0.90	“ ‘”“ likewise”“ ‘’”
5. L3N2798-1💤	0.90	“ ‘”“ “”“ ””
6. L2N71-1💤	0.90	“ ‘”“⏎”“ “”
7. L11N709+1💤	0.90	“⏎”“ ‘”“ “”
8. L16N3206-1💤	0.90	“ “”“⏎”“ ‘”
9. L7N2151-1💤	0.90	“ ‘”“ “”“‘”
10. L3N944-1💤	0.90	“ ‘”“⏎”“ ⏎”

Shadow specimen, apart from the ten: L15N2073-1💤, 0.24.

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 15, neuron 2073, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.69, 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
L15N2073+1 t0-v0.1	0.691	0.67	0.923	1.19	Top boosted tokens ↑
L15N2073-1💤 shadow specimen t0-v0.1	0.547	0.69	0.720	1.29	“ [”“.<”“ noted”“ <”“ referenced”“ [.”“.\”“ now”

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.