INDEX

Explanations

phrases related to actions and events with societal or political implications

oai_token-act-pair · gpt-3.5-turbo

phrases related to accusations and claims of misconduct

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

!'

-0.72

wayne

-0.65

!'"

-0.64

tis

-0.63

NPR

-0.63

gdala

-0.60

();

-0.60

ominated

-0.59

(@

-0.59

onis

-0.58

POSITIVE LOGITS

 unfairly

1.14

"'

1.10

1.01

 unfair

0.89

 insufficient

0.89

 unacceptable

0.89

''

0.87

 "âĢ¦

0.87

"[

0.86

 undue

0.86

Activations Density 0.495%

phrases related to actions and events with societal or political implications

phrases related to accusations and claims of misconduct

No Comments

No Known Activations