INDEX

Explanations

words related to nuclear reactors

oai_token-act-pair · gpt-3.5-turbo

mentions of nuclear reactors

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

 Hills

-0.73

Beg

-0.72

boxing

-0.70

 Barcl

-0.69

CHO

-0.67

Snow

-0.64

 Drawn

-0.64

lihood

-0.64

ãĤĭ

-0.63

haw

-0.63

POSITIVE LOGITS

 reactor

1.36

 reactors

1.29

 dismant

0.83

 coils

0.82

 furnace

0.81

inery

0.81

 plutonium

0.81

 meltdown

0.79

 uranium

0.77

utonium

0.76

Activations Density 0.014%

words related to nuclear reactors

mentions of nuclear reactors

No Comments

No Known Activations