INDEX

Explanations

mentions of locations in various contexts

oai_token-act-pair · gpt-3.5-turbo

instances of commas in the text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

¬¼

-0.74

isi

-0.71

ocl

-0.69

asma

-0.67

enary

-0.66

uber

-0.66

iple

-0.63

sect

-0.62

cel

-0.62

reprene

-0.61

POSITIVE LOGITS

 albeit

1.23

 although

1.22

 whereas

1.14

 though

1.09

 namely

1.04

but

1.03

 including

1.02

 which

0.98

 whereby

0.93

 however

0.92

Activations Density 0.859%

mentions of locations in various contexts

instances of commas in the text

No Comments

No Known Activations