INDEX

Explanations

phrases related to statements or quotes attributed to various individuals

oai_token-act-pair · gpt-3.5-turbo

quotation marks and the phrases contained within them

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.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

•Partition Features Layer 8 GPT2 Small

No Comments

Negative Logits

 Ross

-0.80

 Creator

-0.74

 derby

-0.72

 Story

-0.72

 slate

-0.71

rul

-0.71

 calendar

-0.71

 rental

-0.70

 replacement

-0.69

Yam

-0.69

POSITIVE LOGITS

absolutely

1.66

completely

1.53

probably

1.49

almost

1.47

very

1.46

pretty

1.43

extremely

1.42

clear

1.40

nothing

1.38

significant

1.38

Activations Density 0.139%

phrases related to statements or quotes attributed to various individuals

quotation marks and the phrases contained within them

No Comments

No Known Activations