INDEX

Explanations

descriptions or mentions of documentation

oai_token-act-pair · gpt-3.5-turbo

references to documentation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

otin

-0.82

ellen

-0.78

oter

-0.76

ter

-0.76

cop

-0.75

ofi

-0.75

oker

-0.74

bill

-0.73

ker

-0.72

corn

-0.72

POSITIVE LOGITS

 documentation

1.34

 Documentation

1.12

 docs

0.94

 proofs

0.91

 documenting

0.88

 bindings

0.85

 markup

0.80

 specification

0.80

uments

0.79

 specifications

0.78

Activations Density 0.005%

descriptions or mentions of documentation

references to documentation

No Comments

No Known Activations