INDEX

Explanations

phrases related to ancient Greek characters or symbols

oai_token-act-pair · gpt-3.5-turbo

tokens that are likely technical or scientific symbols and terms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

hower

-0.78

iage

-0.78

ufact

-0.74

bourg

-0.73

management

-0.72

 Brow

-0.71

ilage

-0.69

folk

-0.68

ividual

-0.68

 Lauder

-0.68

POSITIVE LOGITS

Î±

1.82

Î¿

1.80

Î

1.74

ÏĦ

1.70

Î¹

1.65

Ï

1.63

Îº

1.61

á½

1.57

Îµ

1.54

Î½

1.53

Activations Density 0.027%

phrases related to ancient Greek characters or symbols

tokens that are likely technical or scientific symbols and terms

No Comments

No Known Activations