INDEX

Explanations

the word "clone" in various contexts

oai_token-act-pair · gpt-3.5-turbo

references to cloning and clones

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

 Weber

-0.79

 grounding

-0.76

 Wembley

-0.73

MET

-0.72

occup

-0.72

auld

-0.71

Pitt

-0.68

fare

-0.68

 Emanuel

-0.67

 Emirates

-0.66

POSITIVE LOGITS

 clone

3.97

 clones

3.54

 Clone

2.71

clone

2.58

 cloning

2.57

 duplicate

1.25

 hybrids

1.23

copy

1.16

 copied

1.11

 hybrid

1.10

Activations Density 0.031%

the word "clone" in various contexts

references to cloning and clones

No Comments

No Known Activations