INDEX

Explanations

updated information in a document

oai_token-act-pair · gpt-3.5-turbo

sections and headings related to updates and conclusions in documents

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

 jersey

-0.78

kell

-0.75

 Kelly

-0.67

cular

-0.67

cephal

-0.66

ittal

-0.65

raltar

-0.64

 gast

-0.64

ometric

-0.63

 sear

-0.62

POSITIVE LOGITS

~~

1.86

~~~~~~~~

1.80

~~~~~~~~~~~~~~~~

1.78

------------------------------------------------

1.77

~~~~

1.76

********

1.68

======

1.68

 ****

1.65

***

1.65

 -------

1.64

Activations Density 0.113%

updated information in a document

sections and headings related to updates and conclusions in documents

No Comments

No Known Activations