INDEX

Explanations

phrases that involve making distinctions or classifications between different concepts or entities

oai_token-act-pair · gpt-3.5-turbo

the distinction between important and unimportant topics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 radiator

-0.65

 Donation

-0.64

 Indra

-0.61

 Annotations

-0.59

 McDonnell

-0.56

 contacting

-0.56

 Emblem

-0.54

 Sturgeon

-0.54

zzi

-0.53

!:

-0.52

POSITIVE LOGITS

å¤©

0.73

essim

0.69

BILITY

0.64

unch

0.63

idel

0.61

 merely

0.61

olutely

0.60

 ones

0.60

omsday

0.60

ussia

0.59

Activations Density 0.334%

phrases that involve making distinctions or classifications between different concepts or entities

the distinction between important and unimportant topics

No Comments

No Known Activations