INDEX

Explanations

This neuron seems to activate on code snippets, especially declarations and imports, and also news agency names

oai_token-act-pair · gemini-2.0-flash

code strings

np_max-act · gemini-2.0-flash

New Auto-Interp

Configuration

google/gemma-scope-2b-pt-transcoders/layer_12/width_16k/average_l0_6

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.12.ln2.hook_normalized

Architecture

jumprelu_transcoder

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

])))

-0.46

 chung

-0.44

losing

-0.43

EReference

-0.42

fras

-0.42

}}$\\

-0.41

 daarom

-0.40

itys

-0.40

tableFuture

-0.40

 Daarom

-0.39

POSITIVE LOGITS

ThroughAttribute

0.68

FunctionFlags

0.67

rawDesc

0.66

 lenker

0.63

клопе

0.63

 AssemblyCulture

0.61

MigrationBuilder

0.61

 CGRectMake

0.60

ToProps

0.60

yntaxException

0.59

Activations Density 0.139%

This neuron seems to activate on code snippets, especially declarations and imports, and also news agency names

code strings

No Comments

No Known Activations