INDEX

Explanations

the substring "Ind" at various positions within tokens

oai_token-act-pair · gpt-3.5-turbo

occurrences of the term "Ind" followed by various indicators related to classification systems or metrics

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

Not in Any Lists

No Comments

Negative Logits

sonian

-0.86

 compr

-0.79

 litter

-0.75

EStream

-0.73

TION

-0.65

ILA

-0.65

è¦ļéĨĴ

-0.65

Ĥİ

-0.64

 prick

-0.64

 crest

-0.63

POSITIVE LOGITS

irect

1.50

ividually

1.48

icative

1.34

ebted

1.25

icator

1.25

ifferent

1.23

oor

1.21

ications

1.20

icates

1.19

epend

1.17

Activations Density 0.018%

the substring "Ind" at various positions within tokens

occurrences of the term "Ind" followed by various indicators related to classification systems or metrics

No Comments

No Known Activations