INDEX

Explanations

locations or distances mentioned in the text

oai_token-act-pair · gpt-3.5-turbo

phrases indicating distance from a specific location

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

rix

-0.83

unn

-0.81

nell

-0.76

Ĥª

-0.72

rite

-0.71

Article

-0.71

PDF

-0.71

uci

-0.70

quer

-0.67

gow

-0.66

POSITIVE LOGITS

 shore

1.13

 where

1.03

 nearest

0.96

 downtown

0.92

 residences

0.90

 coast

0.87

 headquarters

0.85

 afar

0.85

 home

0.83

 landfall

0.82

Activations Density 0.105%

locations or distances mentioned in the text

phrases indicating distance from a specific location

No Comments

No Known Activations