INDEX

Explanations

terms related to competitive events or races

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

ickr

-0.80

SOURCE

-0.77

reprene

-0.75

 founding

-0.72

leeve

-0.72

etsy

-0.71

anthrop

-0.70

 Anon

-0.67

ulf

-0.67

Library

-0.67

POSITIVE LOGITS

 matchup

1.52

 matchups

1.46

 rematch

1.39

 showdown

1.37

 contest

1.30

Winner

1.18

 contests

1.18

 winner

1.17

stakes

1.15

 playoff

1.14

Activations Density 0.500%

terms related to competitive events or races

No Comments

No Known Activations