INDEX

Explanations

proper names of individuals or organizations

oai_token-act-pair · gpt-3.5-turbo

proper names, particularly those of individuals associated with specific events or statements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

ances

-0.93

ovember

-0.75

ascular

-0.71

omical

-0.70

apple

-0.69

 Indians

-0.64

 Anita

-0.62

 Taiwanese

-0.62

acular

-0.62

 Mortgage

-0.61

POSITIVE LOGITS

ilon

0.87

raltar

0.81

ution

0.77

 Phelps

0.76

ername

0.72

inarily

0.71

minist

0.69

uga

0.68

baugh

0.68

duty

0.67

Activations Density 0.026%

proper names of individuals or organizations

proper names, particularly those of individuals associated with specific events or statements

No Comments

No Known Activations