INDEX

Explanations

words related to playing games or engaging in activities

oai_token-act-pair · gpt-3.5-turbo

instances of the word "play" and its variations in different contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

ortium

-0.83

£ı

-0.67

ixed

-0.65

pora

-0.64

«ĺ

-0.62

gencies

-0.62

thening

-0.62

perial

-0.61

ricane

-0.60

 whisk

-0.59

POSITIVE LOGITS

wright

1.16

plays

1.07

ername

1.05

 play

0.94

 Plays

0.89

play

0.88

Play

0.86

 played

0.86

PLAY

0.79

tested

0.78

Activations Density 0.041%

words related to playing games or engaging in activities

instances of the word "play" and its variations in different contexts

No Comments

No Known Activations