INDEX

Explanations

mentions of organizations, companies, or political entities

oai_token-act-pair · gpt-3.5-turbo

references to political entities and statistics related to budget or production

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

aneous

-0.66

 chronological

-0.65

worldly

-0.63

 unforgettable

-0.61

elfth

-0.60

ranging

-0.60

abilia

-0.59

dden

-0.59

clusive

-0.59

 comprehension

-0.57

POSITIVE LOGITS

 operates

1.44

 employs

1.38

 owns

1.30

 expects

1.30

 conducts

1.28

 publishes

1.27

 invests

1.24

 intends

1.24

 maintains

1.22

has

1.20

Activations Density 0.319%

mentions of organizations, companies, or political entities

references to political entities and statistics related to budget or production

No Comments

No Known Activations