INDEX

Explanations

statements or facts regarded as accurate

oai_token-act-pair · gpt-3.5-turbo

statements related to accuracy and precision

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

loo

-0.80

acid

-0.79

AIN

-0.75

ovan

-0.73

doms

-0.71

Aid

-0.71

inki

-0.71

ATA

-0.70

 Parties

-0.70

joining

-0.69

POSITIVE LOGITS

uracy

1.16

 portrayal

1.00

 inacc

0.97

 inaccurate

0.92

 accuracy

0.92

 depiction

0.89

 appraisal

0.87

imates

0.84

 inaccur

0.83

inately

0.81

Activations Density 0.019%

statements or facts regarded as accurate

statements related to accuracy and precision

No Comments

No Known Activations