INDEX

Explanations

expressions indicating emphasis on a particular point or aspect

oai_token-act-pair · gpt-3.5-turbo

phrases that emphasize the idea of additional information or significance

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

agra

-0.73

itudes

-0.71

plings

-0.71

oes

-0.71

xtap

-0.70

ipment

-0.70

ader

-0.70

uum

-0.69

https

-0.66

xia

-0.65

POSITIVE LOGITS

 important

0.91

 interesting

0.88

 than

0.88

 informative

0.87

 importantly

0.85

 likely

0.83

 challeng

0.83

 worrisome

0.81

 impressive

0.80

 improbable

0.79

Activations Density 0.038%

expressions indicating emphasis on a particular point or aspect

phrases that emphasize the idea of additional information or significance

No Comments

No Known Activations