INDEX

Explanations

phrases related to medical conditions and treatments with occasional mentions of locations or organizations

oai_token-act-pair · gpt-3.5-turbo

instances of numerical data or statistics related to various topics

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

 Mobil

-0.73

Sob

-0.73

 Tunnel

-0.71

 Finish

-0.69

Tru

-0.69

 Soft

-0.69

 scrap

-0.68

 Rally

-0.68

 Camp

-0.68

 Leader

-0.68

POSITIVE LOGITS

Reviewer

1.24

THE

1.09

Anyway

1.09

Secondly

1.07

ÂłÂłÂłÂł

1.06

ÂłÂłÂł

1.05

Contents

1.02

Posted

1.00

BUT

1.00

rawdownloadcloneembedreportprint

1.00

Activations Density 0.410%

phrases related to medical conditions and treatments with occasional mentions of locations or organizations

instances of numerical data or statistics related to various topics

No Known Activations