INDEX

Explanations

words related to technology or commands

oai_token-act-pair · gpt-3.5-turbo

connecting words and punctuation that indicate continuation or emphasis

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

iber

-0.75

assetsadobe

-0.67

pez

-0.64

etts

-0.60

remlin

-0.59

rather

-0.59

ocative

-0.58

artney

-0.58

iru

-0.57

PIN

-0.57

POSITIVE LOGITS

all

1.82

ALL

1.79

all

1.65

All

1.51

All

1.46

 EVERY

1.39

 every

1.36

 everything

1.36

every

1.30

everything

1.27

Activations Density 0.248%

words related to technology or commands

connecting words and punctuation that indicate continuation or emphasis

No Comments

No Known Activations