INDEX

Explanations

phrases related to abilities or actions

oai_token-act-pair · gpt-3.5-turbo

phrases related to limitations and inability

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ghan

-0.64

naire

-0.63

ollah

-0.63

leted

-0.62

rongh

-0.59

 mull

-0.58

rather

-0.57

ortment

-0.57

 rather

-0.56

igl

-0.55

POSITIVE LOGITS

 anymore

1.37

nor

0.95

 anything

0.91

urate

0.86

uate

0.84

any

0.81

 anybody

0.81

 anywhere

0.76

 adequately

0.75

 anyone

0.74

Activations Density 0.200%

phrases related to abilities or actions

phrases related to limitations and inability

No Comments

No Known Activations