INDEX

Explanations

proper nouns, especially related to politics and news

oai_token-act-pair · gpt-3.5-turbo

phrases related to political contexts and figures

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

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

 Sabbath

-1.02

Weiss

-1.02

 Weiss

-1.00

Wisconsin

-0.92

 Somerset

-0.90

 Hess

-0.89

Denver

-0.88

 Hers

-0.85

 Frost

-0.85

 Sheffield

-0.85

POSITIVE LOGITS

 Duterte

2.11

 Filipino

2.03

 Philippine

1.98

uterte

1.98

 Manila

1.86

 Philippines

1.84

 Filip

1.71

 Marcos

1.58

 Rodrigo

1.39

 Manny

1.27

Activations Density 0.288%

proper nouns, especially related to politics and news

phrases related to political contexts and figures

No Comments

No Known Activations