INDEX

Explanations

negative numbers that are related to specific events or metrics

oai_token-act-pair · gpt-3.5-turbo

the end of a document or text segment

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

 prest

-0.64

 destro

-0.64

 Dickinson

-0.63

 caut

-0.59

 Vaugh

-0.56

 proport

-0.54

 Norris

-0.53

 incorpor

-0.51

 Seym

-0.51

kson

-0.51

POSITIVE LOGITS

sized

0.74

mounted

0.73

owned

0.71

level

0.70

backed

0.70

tailed

0.68

initialized

0.68

oriented

0.68

mount

0.67

based

0.66

Activations Density 0.151%

negative numbers that are related to specific events or metrics

the end of a document or text segment

No Comments

No Known Activations