INDEX

Explanations

quotes from individuals

oai_token-act-pair · gpt-3.5-turbo

instances of the word "said."

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

pend

-0.74

paralle

-0.68

 Entered

-0.64

erity

-0.64

 pend

-0.64

 accidents

-0.63

fruit

-0.63

oint

-0.62

blem

-0.62

 manag

-0.61

POSITIVE LOGITS

 sarcast

0.84

 rhet

0.80

 bluntly

0.78

 chuck

0.72

 referring

0.71

rix

0.68

 proudly

0.66

 emphatically

0.64

 said

0.63

 Superintendent

0.63

Activations Density 0.108%

quotes from individuals

instances of the word "said."

No Comments

No Known Activations