INDEX

Explanations

numerical ratios or comparisons

oai_token-act-pair · gpt-3.5-turbo

references to statistical ratios

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

gotten

-0.87

dream

-0.83

inelli

-0.77

uggage

-0.72

REAM

-0.72

 largeDownload

-0.70

pocket

-0.67

âĸº

-0.65

ORE

-0.65

stra

-0.63

POSITIVE LOGITS

xual

1.10

 ratios

0.96

 ratio

0.95

 mismatch

0.88

yip

0.80

 numer

0.78

sum

0.76

yll

0.70

ically

0.69

 imbalance

0.69

Activations Density 0.016%

numerical ratios or comparisons

references to statistical ratios

No Comments

No Known Activations