INDEX

Explanations

phrases related to the type or category of something

oai_token-act-pair · gpt-3.5-turbo

phrases describing types and categories of behaviors or concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

ctors

-0.66

 srfAttach

-0.64

DUP

-0.63

 rece

-0.61

 quar

-0.61

Berry

-0.58

ãĤ¨ãĥ«

-0.57

 Tuls

-0.56

 avail

-0.56

gor

-0.55

POSITIVE LOGITS

isphere

0.79

 course

0.69

ordial

0.67

yle

0.66

illet

0.66

ilion

0.66

uture

0.66

ichick

0.66

worldly

0.66

alien

0.65

Activations Density 0.061%

phrases related to the type or category of something

phrases describing types and categories of behaviors or concepts

No Comments

No Known Activations