INDEX

Explanations

phrases indicating attribution or reporting of statements

oai_token-act-pair · gpt-3.5-turbo

instances of the phrase "to have" as it relates to claims or reported actions

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

catentry

-0.70

etting

-0.68

 Tribune

-0.63

etter

-0.61

 Avalanche

-0.60

 rumored

-0.57

 Dude

-0.57

 Tacoma

-0.57

atin

-0.56

 Crown

-0.56

POSITIVE LOGITS

 been

1.20

been

1.06

 undergone

1.02

 gotten

1.02

 gone

0.96

 taken

0.92

 fared

0.88

gotten

0.87

 Been

0.86

 done

0.85

Activations Density 0.084%

phrases indicating attribution or reporting of statements

instances of the phrase "to have" as it relates to claims or reported actions

No Comments

No Known Activations