Layer 15 · Neuron 6313 · boosted

L15N6313+1

Word quirks:

“ alternative”“ here”“ reflected”“ reflection”“ reflective”“ reflecting”“ this”“ special”

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

Identity (according to the specimen itself)

As an AI assistant developed by Tongyi Lab, I am a large-scale language model created to help users interact with the world in new ways. My core capabilities include:

Core Functionality:

Q&A: I can answer questions about various topics (e.g., science, technology, culture) based on my training data.
Content Creation: I support tasks like writing stories, scripts, or documents. I can also generate code for specific tasks (e.g., Python scripts).
Multilingual Support: I support over 100 languages globally, including but

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. L14N9084+1🕒	0.88	“ reflects”“ reflecting”“ envisioned”
2. L9N5162+1🕒	0.88	“ reflecting”“ reflected”“ reflection”
3. L9N5050-1💤	0.88	“ typical”“ reflecting”“ reflective”
4. L9N1639+1🕒	0.87	“ reflection”“ reflective”“ reflecting”
5. L15N9052-1🕒	0.87	“ reflected”“ reflective”“ reflecting”
6. L14N2265+1💤	0.87	“ here”“ reflected”“ reflecting”
7. L4N3586+1🕒	0.87	“ reflection”“ reflected”“ reflective”
8. L13N6488+1🕒	0.87	“ reflected”“ reflective”“ reflection”
9. L4N6657-1🕒	0.87	“ reflection”“ reflected”“ reflecting”
10. L9N3565+1🕒	0.87	“ reflecting”“ reflected”“ reflection”

Shadow specimen, apart from the ten: L15N6313-1🕒, -0.10.

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 6313, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.47, top-1 agreement 0.73) 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
L15N6313+1 t0-v0.1	0.473	0.73	0.566	1.14	Top boosted tokens ↑
L15N6313-1🕒 shadow specimen t0-v0.1	1.674	0.50	2.016	2.34	“ ‘’”“ “”“ ‘”“ info”“ ””“ certain”“ **“”“ ’”

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.