INDEX

Explanations

phrases related to affirmation or agreement

oai_token-act-pair · gpt-3.5-turbo

repetitive phrases indicating confirmation or acknowledgment

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

aden

-0.71

ULAR

-0.70

è¦ļéĨĴ

-0.69

odium

-0.65

 Mehran

-0.64

ĸļ

-0.63

 bloom

-0.61

uated

-0.61

 pent

-0.60

 cellul

-0.59

POSITIVE LOGITS

eous

1.42

mares

0.93

move

0.93

 Stuff

0.81

ward

0.76

wing

0.76

lords

0.76

lander

0.74

Right

0.73

hand

0.71

Activations Density 0.035%

phrases related to affirmation or agreement

repetitive phrases indicating confirmation or acknowledgment

No Comments

No Known Activations