INDEX

Explanations

instances of the word "name" in various coding contexts

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 0-gemmascope-mlp-65k

Configuration

google/gemma-scope-2b-pt-mlp/layer_0/width_65k/average_l0_72

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.0.hook_mlp_out

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 pleaſure

-0.86

 reaſon

-0.80

 myſelf

-0.80

 fhort

-0.78

 anſ

-0.77

 fhew

-0.76

 uſe

-0.75

 houſe

-0.75

 Houſe

-0.75

 Anſ

-0.74

POSITIVE LOGITS

 estekak

0.61

['

0.60

发表于

0.57

Życiorys

0.56

]!='

0.56

["

0.55

WriteLiteral

0.55

 跳转至

0.53

HttpFoundation

0.53

des

0.52

Activations Density 0.016%

instances of the word "name" in various coding contexts

No Comments

No Known Activations