INDEX

Explanations

phrases related to looking at something in its entirety or as a complete entity

oai_token-act-pair · gpt-3.5-turbo

references to collective entities or groups as a single unit

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

minist

-0.89

ufact

-0.80

yrinth

-0.71

 Machines

-0.69

CHAT

-0.65

 Specialist

-0.64

Fax

-0.63

Yak

-0.63

Tuc

-0.61

JJ

-0.61

POSITIVE LOGITS

heartedly

1.06

hearted

0.93

meal

0.70

,...

0.67

sheet

0.62

 stride

0.61

orse

0.58

âĶĢâĶĢâĶĢâĶĢâĶĢâĶĢâĶĢâĶĢ

0.58

cule

0.58

 spectrum

0.58

Activations Density 0.018%

phrases related to looking at something in its entirety or as a complete entity

references to collective entities or groups as a single unit

No Comments

No Known Activations