INDEX

Explanations

terms related to political or regulatory actions

oai_token-act-pair · gpt-3.5-turbo

words related to deregulation and related concepts

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

uyomi

-0.89

ĸļ

-0.86

DragonMagazine

-0.85

 Flavoring

-0.81

 Wage

-0.73

 Showdown

-0.71

ocene

-0.70

 Opportun

-0.69

ORY

-0.68

aceous

-0.68

POSITIVE LOGITS

arest

1.16

 Blasio

1.13

activate

1.07

utsche

1.05

activated

1.05

cember

1.02

classified

1.00

jected

0.99

uter

0.98

 facto

0.97

Activations Density 0.027%

terms related to political or regulatory actions

words related to deregulation and related concepts

No Comments

No Known Activations