INDEX

Explanations

words related to actions or processes that continue indefinitely

oai_token-act-pair · gpt-3.5-turbo

various forms of words that end in "end" or "ending"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

hee

-0.74

ushing

-0.71

iard

-0.70

 Selling

-0.66

apult

-0.65

asses

-0.64

heast

-0.63

shop

-0.63

hing

-0.63

-0.62

POSITIVE LOGITS

lich

0.78

uality

0.78

ulum

0.75

lihood

0.73

ught

0.70

ragon

0.68

rils

0.65

adoes

0.63

iously

0.63

encies

0.61

Activations Density 0.035%

words related to actions or processes that continue indefinitely

various forms of words that end in "end" or "ending"

No Comments

No Known Activations