INDEX

Explanations

comparative statements emphasizing the degree of positivity or negativity

oai_token-act-pair · gpt-3.5-turbo

comparative evaluations and degrees of quality regarding various subjects

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

arnaev

-0.70

hao

-0.66

ping

-0.64

rug

-0.64

 brothers

-0.64

 Romans

-0.60

azor

-0.58

enium

-0.58

 Corm

-0.58

aternity

-0.58

POSITIVE LOGITS

é¾įå

0.83

åħī

0.79

 glance

0.79

FontSize

0.75

ç¥ŀ

0.72

 BaseType

0.70

é¾įå¥ĳå£«

0.70

 è£ıè

0.69

ICA

0.69

ãĥīãĥ©ãĤ´ãĥ³

0.69

Activations Density 0.035%

comparative statements emphasizing the degree of positivity or negativity

comparative evaluations and degrees of quality regarding various subjects

No Comments

No Known Activations