INDEX

Explanations

mention symbolizing strength or official information

oai_token-act-pair · gpt-3.5-turbo

suffixes associated with specialized or technical terms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.6.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

SIZE

-0.73

ers

-0.70

ccording

-0.68

erest

-0.66

ership

-0.65

ering

-0.65

ERC

-0.63

hovah

-0.63

sterdam

-0.63

ijing

-0.62

POSITIVE LOGITS

ror

0.98

oute

0.95

rors

0.93

aton

0.91

iffe

0.88

rane

0.87

idge

0.86

 extraord

0.82

jee

0.82

 Than

0.81

Activations Density 0.131%

mention symbolizing strength or official information

suffixes associated with specialized or technical terms

No Comments

No Known Activations