INDEX

Explanations

characters that make up a specific symbol or special character in text

oai_token-act-pair · gpt-3.5-turbo

instances of formatted text or unusual text patterns

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 Alexa

-0.73

uko

-0.68

 Misty

-0.65

sil

-0.59

Ã±a

-0.58

sburg

-0.58

 spirit

-0.57

 Sense

-0.57

isd

-0.56

 Aval

-0.55

POSITIVE LOGITS

à¼

0.91

IBLE

0.87

?????-?????-

0.87

Prosecut

0.81

PART

0.78

Anonymous

0.76

Christ

0.74

Wait

0.74

Marco

0.74

WARNING

0.74

Activations Density 0.109%

characters that make up a specific symbol or special character in text

instances of formatted text or unusual text patterns

No Comments

No Known Activations