INDEX

Explanations

terms relating to newness or innovation

oai_token-act-pair · gpt-3.5-turbo

references to "new" concepts, ideas, or entities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

ality

-0.70

LET

-0.66

ALT

-0.63

zzi

-0.61

APD

-0.60

verbal

-0.60

NP

-0.60

endez

-0.59

oids

-0.58

hunt

-0.56

POSITIVE LOGITS

new

3.03

new

2.31

 newly

1.99

 newest

1.78

 newer

1.68

 newfound

1.67

NEW

1.64

 newcomers

1.52

 fresh

1.43

 revamped

1.36

Activations Density 0.095%

terms relating to newness or innovation

references to "new" concepts, ideas, or entities

No Comments

No Known Activations