INDEX

Explanations

proper nouns, particularly names of companies or entities

oai_token-act-pair · gpt-3.5-turbo

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

rawdownloadcloneembedreportprint

-0.75

bleacher

-0.71

ĸļ

-0.71

uyomi

-0.67

EStreamFrame

-0.67

natureconservancy

-0.63

 Bundy

-0.63

tnc

-0.62

ļéĨĴ

-0.62

Sweet

-0.61

POSITIVE LOGITS

·

0.76

Ħ

0.74

ĩ

0.73

«

0.70

LM

0.65

ĺ

0.65

¹

0.61

ifa

0.60

oids

0.60

ľ

0.59

Activations Density 0.481%

proper nouns, particularly names of companies or entities

No Comments

No Known Activations