INDEX

Explanations

contractions and possessive forms

oai_token-act-pair · gpt-3.5-turbo

phrases indicating ongoing states or situations

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

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

 newsp

-0.65

naires

-0.65

Queue

-0.64

ench

-0.63

-0.62

illac

-0.62

luaj

-0.60

ãĤµ

-0.59

 subscrib

-0.58

IFA

-0.58

POSITIVE LOGITS

 doubtful

1.20

 unclear

1.16

 unlikely

1.12

 still

1.07

 undeniable

1.07

 nevertheless

1.04

 nonetheless

1.04

 clear

1.01

 worth

0.99

 hardly

0.96

Activations Density 0.126%

contractions and possessive forms

phrases indicating ongoing states or situations

No Comments

No Known Activations