INDEX

Explanations

words ending in '-en' and '-ed'

oai_token-act-pair · gpt-3.5-turbo

various forms and conjugations of the prefix "en-"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

apest

-0.78

yip

-0.71

hops

-0.66

jri

-0.65

appropriate

-0.64

 incon

-0.64

cules

-0.63

 kittens

-0.61

 heaviest

-0.60

 crop

-0.60

POSITIVE LOGITS

ciating

1.08

enment

0.97

ãĤ¨ãĥ«

0.91

yll

0.83

à¼

0.76

achment

0.75

oyer

0.74

emies

0.70

vironment

0.69

ment

0.66

Activations Density 0.068%

words ending in '-en' and '-ed'

various forms and conjugations of the prefix "en-"

No Comments

No Known Activations