INDEX

Explanations

proper nouns or entities mentioned in text, especially associated with identity or labeling

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "Name" in various contexts

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

 relaxation

-0.68

rals

-0.65

aris

-0.65

rats

-0.64

 stim

-0.63

 Revel

-0.63

urable

-0.62

trap

-0.61

stim

-0.61

 screenings

-0.61

POSITIVE LOGITS

 Name

3.79

Name

2.64

 NAME

2.52

 Names

2.43

 name

2.31

name

2.02

Names

1.89

 names

1.86

NAME

1.69

 surname

1.53

Activations Density 0.010%

proper nouns or entities mentioned in text, especially associated with identity or labeling

occurrences of the word "Name" in various contexts

No Comments

No Known Activations