INDEX

Explanations

phrases indicating the start of a list or a new topic, specifically starting with words like "Some" or "Like"

oai_token-act-pair · gpt-3.5-turbo

instances of the word "some" and its variations in the context of discussions

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

eret

-0.65

runtime

-0.63

BuyableInstoreAndOnline

-0.61

ifice

-0.61

âĢ¢âĢ¢âĢ¢âĢ¢

-0.60

ACTED

-0.60

 theirs

-0.58

abad

-0.58

âĦ¢:

-0.58

RY

-0.58

POSITIVE LOGITS

how

1.43

body

1.29

 argue

1.03

 speculate

0.96

 commentators

0.92

ones

0.90

 observers

0.90

place

0.89

 commenters

0.87

 people

0.85

Activations Density 0.077%

phrases indicating the start of a list or a new topic, specifically starting with words like "Some" or "Like"

instances of the word "some" and its variations in the context of discussions

No Comments

No Known Activations