INDEX

Explanations

numerical values mentioned in the text

oai_token-act-pair · gpt-3.5-turbo

instances of specific identifiers and terms associated with measurements or numerical values

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

 minds

-0.69

 Guth

-0.67

aired

-0.66

 mand

-0.66

 paradise

-0.65

club

-0.64

 laps

-0.62

 rapt

-0.61

 impro

-0.61

 grou

-0.60

POSITIVE LOGITS

iewicz

2.63

ints

2.58

cci

1.94

ÅĤ

1.19

ulz

1.18

zzi

1.16

cius

0.99

fc

0.99

ixel

0.95

coni

0.92

Activations Density 0.071%

numerical values mentioned in the text

instances of specific identifiers and terms associated with measurements or numerical values

No Comments

No Known Activations