Specimen ledger / L7N4616-1

Layer 7 · Neuron 4616 · suppressed

L7N4616-1

Word quirks:

“ representing”“ representations”“ representation”“ referenced”“ generating”“ metaphor”“ **“”“ acknowledgment”

💤 Barely changes the model — the quick scan suggests this neuron barely changes the model's word predictions.

Identity (according to the specimen itself)

Hello! I am Qwen3.5, a large language model developed by Alibaba Cloud. My identity is defined as follows:

My Identity:

I am a large language model with a text-generation capability, trained on high-quality data to understand and generate human language.

My Capabilities:

I can understand complex queries in natural language, whether they involve questions about facts, reasoning tasks, or creative prompts.
I support multilingual dialogue, allowing me to communicate fluently across many languages.
I can perform mathematical computations and logical reasoning to solve problems

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. L5N3638+1🕒	0.85	“ representation”“ ‘”“ representing”
2. L7N3818+1💤	0.84	“ representation”“ representing”“ representations”
3. L15N5974+1🕒	0.84	“ representation”“ metaphor”“ represent”
4. L10N7854-1🕒	0.84	“ representing”“ representation”“ reflecting”
5. L9N2864-1🕒	0.83	“ representing”“ represented”“ representation”
6. L6N1634+1🕒	0.83	“ representation”“ reference”“ representations”
7. L5N3183-1🕒	0.83	“ referenced”“ reference”“ representation”
8. L10N2490-1🕒	0.82	“ representation”“ representative”“ representing”
9. L5N2845-1🕒	0.82	“ representation”“<\|im_end\|>”“ metaphor”
10. L10N7062-1🕒	0.81	“ representation”“ representations”“ representing”

Shadow specimen, apart from the ten: L7N4616+1🕒, 0.05.

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 4616, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.95, top-1 agreement 0.63) 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
L7N4616+1🕒 shadow specimen t0-v0.1	1.625	0.54	1.999	1.68	“ experiment”“ experiments”“ experimental”“ scientific”“ biological”“ biology”“ scientist”“ research”
L7N4616-1 t0-v0.1	0.949	0.63	1.197	1.47	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.