INDEX

Explanations

items listed using the word "lists."

oai_token-act-pair · gpt-3.5-turbo

references to lists

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

amen

-0.74

acea

-0.67

 srfAttach

-0.66

ruption

-0.64

cling

-0.63

 Skydragon

-0.62

asin

-0.62

ifice

-0.61

cession

-0.61

alty

-0.60

POSITIVE LOGITS

 lists

3.94

 Lists

3.02

 list

2.46

lists

2.46

 List

1.94

list

1.81

List

1.67

 LIST

1.66

 listings

1.58

 rosters

1.51

Activations Density 0.013%

items listed using the word "lists."

references to lists

No Comments

No Known Activations