INDEX

Explanations

phrases related to individuals' names

oai_token-act-pair · gpt-3.5-turbo

references to specific individuals mentioned 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 @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

orthy

-0.78

vation

-0.75

ctica

-0.74

 Sussex

-0.74

phys

-0.70

itably

-0.70

ilight

-0.69

aciously

-0.69

itable

-0.68

phia

-0.66

POSITIVE LOGITS

enegger

0.78

ertodd

0.78

 reluct

0.74

ham

0.71

 Pengu

0.71

mann

0.70

ensen

0.70

linger

0.69

 Clicker

0.66

aturdays

0.65

Activations Density 0.087%

phrases related to individuals' names

references to specific individuals mentioned in the text

No Comments

No Known Activations