INDEX

Explanations

the word "of" used in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.8.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

 agre

-0.83

ĸļ

-0.81

gow

-0.79

omnia

-0.78

 unfocusedRange

-0.75

acus

-0.74

urden

-0.73

Rated

-0.72

ration

-0.72

rated

-0.70

POSITIVE LOGITS

 dozen

1.29

 hundred

1.26

 weeks

1.02

 thousand

0.97

 paragraphs

0.93

 summers

0.91

 thirds

0.91

 tablespoons

0.89

 beers

0.87

 clicks

0.86

Activations Density 0.056%

the word "of" used in various contexts

No Comments

No Known Activations