INDEX

Explanations

replies and responses to comments in online forums or social media

oai_token-act-pair · gpt-3.5-turbo

instances of the word "Reply" in various contexts related to online discussions or comments

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

 roller

-0.66

iders

-0.63

 Hornets

-0.63

 Newport

-0.63

Nim

-0.63

 Bett

-0.63

 Leap

-0.62

 Raleigh

-0.62

Bav

-0.61

inelli

-0.60

POSITIVE LOGITS

reply

1.20

 reply

1.06

ysis

1.02

 replies

0.99

 Delete

0.89

imei

0.89

 answ

0.87

swer

0.86

 thereto

0.85

isson

0.83

Activations Density 0.020%

replies and responses to comments in online forums or social media

instances of the word "Reply" in various contexts related to online discussions or comments

No Comments

No Known Activations