INDEX

Explanations

proper names and dates formatted in a news article style

oai_token-act-pair · gpt-3.5-turbo

references to news articles or reports

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

ĪĴ

-0.78

 brakes

-0.70

worms

-0.65

onics

-0.65

 verbs

-0.64

 prol

-0.64

 favor

-0.63

 pets

-0.63

 leagues

-0.62

 cush

-0.62

POSITIVE LOGITS

Jonathan

0.89

David

0.88

Joshua

0.87

File

0.87

String

0.85

Jason

0.84

Carl

0.83

Jim

0.83

Scott

0.83

Leon

0.82

Activations Density 0.046%

proper names and dates formatted in a news article style

references to news articles or reports

No Comments

No Known Activations