INDEX

Explanations

phrases related to prompts or requests

oai_token-act-pair · gpt-3.5-turbo

calls to action or directives

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

oin

-0.61

 prophes

-0.61

 pseudonym

-0.59

 appearance

-0.57

è¦ļéĨĴ

-0.57

tom

-0.57

 makeshift

-0.56

 evoke

-0.56

ISC

-0.55

iege

-0.55

POSITIVE LOGITS

TING

1.14

rid

0.96

aways

0.90

away

0.89

terson

0.79

zl

0.78

 Started

0.75

 Tickets

0.74

gets

0.72

gotten

0.70

Activations Density 0.037%

phrases related to prompts or requests

calls to action or directives

No Comments

No Known Activations