INDEX

Explanations

words related to actions in the past and present

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

past tense action verbs ending in 'ed' or 'd'

oai_token-act-pair · gpt-3.5-turbo

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

cipled

-0.81

CLUD

-0.63

bent

-0.62

ussed

-0.61

oided

-0.59

 Entered

-0.59

 healed

-0.58

board

-0.57

cised

-0.56

 Provided

-0.55

POSITIVE LOGITS

pez

1.03

zed

0.82

omsday

0.81

oms

0.76

zing

0.75

lege

0.74

aughters

0.69

xy

0.68

ettings

0.67

berman

0.66

Activations Density 0.056%

words related to actions in the past and present

past tense action verbs ending in 'ed' or 'd'

No Comments

No Known Activations