INDEX

Explanations

phrases related to prestigious awards such as Nobel Prize and Pulitzer Prize

oai_token-act-pair · gpt-3.5-turbo

mentions of prestigious awards, particularly the Nobel and Pulitzer Prizes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

addock

-0.73

icago

-0.72

alach

-0.71

aves

-0.71

aved

-0.67

andering

-0.66

 Reyn

-0.65

afort

-0.65

\-

-0.65

othy

-0.64

POSITIVE LOGITS

 Prize

1.20

 laureate

1.13

 laure

1.13

 Nobel

1.12

 Laure

1.00

Nob

0.92

 prize

0.78

 medal

0.77

 Lect

0.76

 physicist

0.74

Activations Density 0.006%

phrases related to prestigious awards such as Nobel Prize and Pulitzer Prize

mentions of prestigious awards, particularly the Nobel and Pulitzer Prizes

No Comments

No Known Activations