INDEX

Explanations

phrases related to positive evaluations and impactful actions

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate improvement or upward movement

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

lass

-0.66

wives

-0.65

sw

-0.63

bats

-0.62

 Registry

-0.61

 Ascension

-0.61

FG

-0.60

Shift

-0.60

gins

-0.59

 reservations

-0.59

POSITIVE LOGITS

Downloadha

0.82

ometimes

0.80

everal

0.74

irtual

0.74

muc

0.67

hift

0.66

 scaven

0.66

udeb

0.65

inger

0.63

omething

0.63

Activations Density 0.928%

phrases related to positive evaluations and impactful actions

phrases that indicate improvement or upward movement

No Comments

No Known Activations