INDEX

Explanations

opinions or beliefs that are validated or confirmed

oai_token-act-pair · gpt-3.5-turbo

statements affirming correctness or truth

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 Remastered

-0.81

gins

-0.78

 Flavoring

-0.73

 scrim

-0.66

effic

-0.64

 Pastebin

-0.64

hens

-0.63

 Qiao

-0.63

 Pharma

-0.61

 Gong

-0.61

POSITIVE LOGITS

footed

1.09

eous

0.96

headed

0.95

 about

0.90

 wing

0.79

wing

0.78

fully

0.76

eyed

0.74

about

0.73

 About

0.72

Activations Density 0.027%

opinions or beliefs that are validated or confirmed

statements affirming correctness or truth

No Comments

No Known Activations