INDEX

Explanations

numerical values and different sections in a document

oai_token-act-pair · gpt-3.5-turbo

phrases and questions that indicate uncertainty or seek clarification

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

eri

-0.65

sen

-0.62

uma

-0.60

uct

-0.58

opus

-0.58

anni

-0.57

 Revision

-0.57

ctory

-0.57

itely

-0.56

issance

-0.56

POSITIVE LOGITS

 where

2.07

where

2.02

 Where

1.77

Where

1.67

 WHERE

1.64

 wherein

1.48

WHERE

1.48

 whence

1.43

 Places

1.22

 wherever

1.15

Activations Density 0.134%

numerical values and different sections in a document

phrases and questions that indicate uncertainty or seek clarification

No Comments

No Known Activations