INDEX

Explanations

phrases related to security or potential threats

oai_token-act-pair · gpt-3.5-turbo

references to bouncers and notable individuals or terms associated with performances or events

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

ocene

-0.82

ories

-0.81

yrus

-0.81

orers

-0.79

iance

-0.78

aeda

-0.77

ocalypse

-0.73

sym

-0.73

heny

-0.73

algia

-0.71

POSITIVE LOGITS

âĸ¬

0.94

cÃ©

0.85

pload

0.82

 lineback

0.76

nen

0.76

cin

0.74

cel

0.73

umber

0.72

ciples

0.70

cer

0.70

Activations Density 0.029%

phrases related to security or potential threats

references to bouncers and notable individuals or terms associated with performances or events

No Comments

No Known Activations