INDEX

Explanations

Roman numerals related to volume or editions of a series

oai_token-act-pair · gpt-3.5-turbo

references to specific volumes or editions of a series or collection

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

rules

-0.79

 value

-0.74

boards

-0.70

original

-0.70

blocks

-0.70

winner

-0.68

block

-0.66

prises

-0.64

value

-0.62

giving

-0.61

POSITIVE LOGITS

VI

1.09

III

1.01

VI

0.99

ISION

0.90

BuyableInstoreAndOnline

0.88

urious

0.87

VII

0.83

XX

0.82

KING

0.81

engeance

0.80

Activations Density 0.006%

Roman numerals related to volume or editions of a series

references to specific volumes or editions of a series or collection

No Comments

No Known Activations