INDEX

Explanations

instances of locations, particularly focusing on mentions of the Philippines

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ãĥł

-0.75

ership

-0.75

eer

-0.73

rians

-0.72

 Macintosh

-0.72

 Bauer

-0.70

 Ãĸ

-0.70

jee

-0.69

heimer

-0.68

 Cobb

-0.67

POSITIVE LOGITS

pes

0.98

 Islands

0.96

 Duterte

0.93

 typh

0.92

 Philippines

0.87

 Manila

0.86

uterte

0.85

 Rodrigo

0.82

uala

0.79

 Philippine

0.77

Activations Density 0.049%

instances of locations, particularly focusing on mentions of the Philippines

No Comments

No Known Activations