INDEX

Explanations

capital letter combinations that likely refer to specific functions or variables in a programming context

oai_token-act-pair · gpt-3.5-turbo

references to specific identifiers or codes within a structured data format

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

 polio

-0.73

EPA

-0.72

GMO

-0.70

 warrants

-0.67

 prosecuted

-0.67

 perme

-0.66

 Guang

-0.66

DUI

-0.66

 stressed

-0.65

 Mits

-0.65

POSITIVE LOGITS

ibrary

1.60

ocket

1.57

oad

1.52

olor

1.51

arget

1.51

essage

1.50

ound

1.50

ustom

1.50

ovie

1.49

ategory

1.49

Activations Density 0.233%

capital letter combinations that likely refer to specific functions or variables in a programming context

references to specific identifiers or codes within a structured data format

No Comments

No Known Activations