INDEX

Explanations

references to the most recent/latest events or news

oai_token-act-pair · gpt-3.5-turbo

phrases referencing "latest" news or updates

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

iac

-0.68

ality

-0.67

ias

-0.63

iva

-0.62

iator

-0.62

ier

-0.61

less

-0.61

lessness

-0.59

lot

-0.59

 representing

-0.58

POSITIVE LOGITS

 latest

3.66

 newest

2.52

latest

2.30

 Latest

2.18

Latest

1.83

 hottest

1.38

 recent

1.36

 oldest

1.25

 earliest

1.24

 upcoming

1.23

Activations Density 0.021%

references to the most recent/latest events or news

phrases referencing "latest" news or updates

No Comments

No Known Activations