INDEX

Explanations

proper nouns related to politics and legislation

oai_token-act-pair · gpt-3.5-turbo

references to government acts and specific officials or titles

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

uggest

-0.91

sonian

-0.90

ovie

-0.81

itz

-0.80

auts

-0.75

ennett

-0.73

essa

-0.72

 Jericho

-0.71

ettings

-0.71

irk

-0.70

POSITIVE LOGITS

à¤

1.57

 à¤

1.22

à¥

1.08

ħ

0.86

à¤¾

0.85

¨

0.85

¹

0.83

à¦

0.83

¤

0.82

ãĥ©

0.82

Activations Density 0.003%

proper nouns related to politics and legislation

references to government acts and specific officials or titles

No Comments

No Known Activations