INDEX

Explanations

review of abstracts

np_acts-logits-general · gemini-2.5-flash-lite

The neuron strongly activates on common English function words (articles, conjunctions, prepositions) and punctuation.

oai_token-act-pair · o4-mini Triggered by @jyhe0408

This neuron specializes in spotting academic‐style paper titles or headings, especially the “X of a/for Y” and “Development/Review of…” constructions typical of scientific abstracts.

oai_token-act-pair · o4-mini Triggered by @jyhe0408

New Auto-Interp

Configuration

google/gemma-scope-27b-pt-res/layer_10/width_131k

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 wasn

-1.28

')

-1.28

 there

-1.22

}\}$

-1.22

 because

-1.20

")

-1.19

?」

-1.18

 only

-1.18

'])

-1.16

 every

-1.15

POSITIVE LOGITS

of

1.24

how

1.13

后面

1.12

 Suomessa

1.05

そうに

1.03

地看着

1.02

目で

1.02

 знамени

0.99

noma

0.98

옐

0.96

Activations Density 0.575%

review of abstracts

The neuron strongly activates on common English function words (articles, conjunctions, prepositions) and punctuation.

This neuron specializes in spotting academic‐style paper titles or headings, especially the “X of a/for Y” and “Development/Review of…” constructions typical of scientific abstracts.

No Comments

No Known Activations

review of abstracts

The neuron strongly activates on common English function words (articles, conjunctions, prepositions) and punctuation.

This neuron specializes in spotting academic‐style paper titles or headings, especially the “X of a/for Y” and “Development/Review of…” constructions typical of scientific abstracts.

No Comments

No Known Activations