INDEX

Explanations

phrases related to exclusion or prohibition

oai_token-act-pair · gpt-3.5-turbo

phrases related to exclusion or staying away from certain people, places, or situations

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

elf

-1.05

ELS

-0.71

IZE

-0.71

ilib

-0.65

olars

-0.63

oria

-0.62

Scene

-0.61

 antioxid

-0.61

iov

-0.61

ternity

-0.60

POSITIVE LOGITS

nit

0.80

IPM

0.66

ned

0.66

 bother

0.66

 Avenger

0.64

0.63

 ãĤµ

0.63

vana

0.63

 altogether

0.63

 Pryor

0.63

Activations Density 0.051%

phrases related to exclusion or prohibition

phrases related to exclusion or staying away from certain people, places, or situations

No Comments

No Known Activations