INDEX

Explanations

positive descriptions or acknowledgments of accomplishments

oai_token-act-pair · gpt-3.5-turbo

phrases indicating accomplishment or positive evaluation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

pora

-0.71

thur

-0.64

soDeliveryDate

-0.62

jay

-0.62

iframe

-0.61

TN

-0.60

devices

-0.60

letters

-0.59

 passport

-0.59

 Issue

-0.57

POSITIVE LOGITS

nered

0.94

itud

0.86

rand

0.73

structed

0.72

dden

0.69

kered

0.68

alian

0.67

umenthal

0.64

ows

0.63

stood

0.63

Activations Density 0.113%

positive descriptions or acknowledgments of accomplishments

phrases indicating accomplishment or positive evaluation

No Comments

No Known Activations