INDEX

Explanations

educational, technical, or medical terms

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

programming-related identifiers and class/variable names.

oai_token-act-pair · claude-3-7-sonnet-20250219 Triggered by @neilrathi

This neuron responds to programming-style identifiers or proper names (e.g. class, table, or field names) rather than ordinary words.

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

 cathédrale

-1.39

 proberen

-1.36

 眼睛

-1.33

使用了

-1.32

 decisões

-1.30

 conseguido

-1.30

 dificuldades

-1.29

是我们

-1.27

必须要

-1.23

mitir

-1.23

POSITIVE LOGITS

1.87

 nawet

1.35

by

1.34

omalies

1.28

 these

1.27

↵

1.27

 memutus

1.25

tismo

1.23

乙女

1.23

 spécialisé

1.23

Activations Density 0.009%

educational, technical, or medical terms

programming-related identifiers and class/variable names.

This neuron responds to programming-style identifiers or proper names (e.g. class, table, or field names) rather than ordinary words.

No Comments

No Known Activations

educational, technical, or medical terms

programming-related identifiers and class/variable names.

This neuron responds to programming-style identifiers or proper names (e.g. class, table, or field names) rather than ordinary words.

No Comments

No Known Activations