INDEX

Explanations

instances where an alternative or contrasting action is taken

oai_token-act-pair · gpt-3.5-turbo

the word "Instead" indicating a contrast or alternative statement

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

emate

-0.60

 compr

-0.55

 Otherwise

-0.55

"},"

-0.55

MpServer

-0.52

ental

-0.51

AV

-0.48

gin

-0.48

OTHER

-0.47

absolutely

-0.47

POSITIVE LOGITS

0.82

of

0.78

 thereof

0.75

,.

0.69

,,

0.67

 opting

0.67

,...

0.66

!,

0.64

,-

0.62

we

0.62

Activations Density 0.052%

instances where an alternative or contrasting action is taken

the word "Instead" indicating a contrast or alternative statement

No Comments

No Known Activations