INDEX

Explanations

dates and events mentioned within synopses or titles

oai_token-act-pair · gpt-3.5-turbo

elements of media formatting, such as episode titles and air dates

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 someday

-0.67

 existential

-0.63

 coerc

-0.54

onym

-0.53

 squee

-0.53

 arthritis

-0.52

 allergies

-0.52

 smugg

-0.52

 dementia

-0.52

 altru

-0.52

POSITIVE LOGITS

 consisted

0.71

 consisting

0.65

mediately

0.63

 commenced

0.63

 comprising

0.62

 wherein

0.61

undown

0.58

aired

0.58

xtap

0.58

adjusted

0.57

Activations Density 0.857%

dates and events mentioned within synopses or titles

elements of media formatting, such as episode titles and air dates

No Comments

No Known Activations