INDEX

Explanations

phrases related to data and evidence

oai_token-act-pair · gpt-3.5-turbo

references to evidence and information in research or studies

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

inia

-0.76

merce

-0.62

ptives

-0.59

DonaldTrump

-0.58

ategor

-0.58

 Reload

-0.57

 Frey

-0.56

 Disorder

-0.55

Purchase

-0.55

 Landing

-0.54

POSITIVE LOGITS

iest

1.31

 surrounding

0.84

liest

0.80

fulness

0.79

 underpin

0.75

 behind

0.75

isphere

0.73

 continuum

0.72

 aspect

0.72

 spectrum

0.71

Activations Density 0.258%

phrases related to data and evidence

references to evidence and information in research or studies

No Comments

No Known Activations