INDEX

Explanations

statements of opinions or reported information

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate reporting or statements attributed to sources

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Written

-0.76

ĸļå£«

-0.76

 conflic

-0.73

ÃĥÃĤÃĥÃĤ

-0.73

 earthqu

-0.71

 pione

-0.70

otype

-0.69

respective

-0.69

thia

-0.69

theless

-0.68

POSITIVE LOGITS

 goodbye

1.03

olate

0.73

 they

0.68

MSG

0.68

ysis

0.68

ings

0.67

NAD

0.66

ofi

0.64

 hello

0.63

 Goodbye

0.63

Activations Density 0.086%

statements of opinions or reported information

phrases that indicate reporting or statements attributed to sources

No Comments

No Known Activations