INDEX

Explanations

phrases related to locations or events

oai_token-act-pair · gpt-3.5-turbo

references to the season of fall

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

>>>>>>>>

-0.76

cffff

-0.74

llular

-0.71

ittees

-0.68

 appealed

-0.66

PLA

-0.66

":[{"

-0.63

kefeller

-0.63

iliary

-0.63

acts

-0.62

POSITIVE LOGITS

ujah

1.18

opian

1.09

acies

0.95

acy

0.90

mary

0.87

acious

0.86

brook

0.86

ible

0.79

Fall

0.77

ibility

0.76

Activations Density 0.018%

phrases related to locations or events

references to the season of fall

No Comments

No Known Activations