INDEX

Explanations

technology-related terms or product specifications

oai_token-act-pair · gpt-3.5-turbo

instances of commas in a list format

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

¬¼

-0.75

cule

-0.75

endo

-0.74

izen

-0.68

ocalypse

-0.67

½

-0.65

rio

-0.64

oll

-0.62

sylv

-0.61

worldly

-0.61

POSITIVE LOGITS

 which

1.07

 whereby

1.01

 whereas

0.98

 albeit

0.93

 although

0.91

 wherein

0.83

 indicating

0.83

 namely

0.83

which

0.82

 consisting

0.81

Activations Density 0.484%

technology-related terms or product specifications

instances of commas in a list format

No Comments

No Known Activations