INDEX

Explanations

mentions of smaller entities or the concept of something being smaller in comparison

oai_token-act-pair · gpt-3.5-turbo

references to smaller entities or groups

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

ILLE

-0.78

onso

-0.74

OOL

-0.70

YR

-0.70

SEA

-0.70

RA

-0.69

plain

-0.69

Maps

-0.66

shire

-0.65

Wikipedia

-0.64

POSITIVE LOGITS

 sized

1.03

 intestine

1.00

 than

1.00

 fractions

0.91

pox

0.89

 batches

0.88

 diameter

0.86

 versions

0.84

thumbnails

0.84

 Than

0.83

Activations Density 0.016%

mentions of smaller entities or the concept of something being smaller in comparison

references to smaller entities or groups

No Comments

No Known Activations