INDEX

Explanations

phrases indicating continuation or progression of a story or discussion

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "so" used to introduce statements or comments

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

SourceFile

-0.70

PF

-0.65

DERR

-0.62

Angelo

-0.60

ļéĨĴ

-0.60

rule

-0.60

verage

-0.59

ailability

-0.59

zag

-0.57

saf

-0.57

POSITIVE LOGITS

 yeah

1.14

uh

0.91

yes

0.86

 alas

0.85

um

0.84

 needless

0.83

 beware

0.81

 congr

0.81

 please

0.80

 unsurprisingly

0.80

Activations Density 0.068%

phrases indicating continuation or progression of a story or discussion

occurrences of the word "so" used to introduce statements or comments

No Comments

No Known Activations