INDEX

Explanations

locations such as villages, towns, districts, states, and cities mentioned in a document, particularly names of specific regions

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Weiss

-0.85

strom

-0.76

Cub

-0.76

Blake

-0.75

angelo

-0.74

BILITIES

-0.70

HEAD

-0.68

sonian

-0.68

scenes

-0.67

iannopoulos

-0.67

POSITIVE LOGITS

 Delhi

1.68

 Mumbai

1.64

 Kerala

1.62

 Maharashtra

1.62

 Gujarat

1.59

 Bihar

1.59

 Chennai

1.59

 Hyder

1.58

 Bangalore

1.57

 Uttar

1.56

Activations Density 0.220%

locations such as villages, towns, districts, states, and cities mentioned in a document, particularly names of specific regions

No Comments

No Known Activations