INDEX

Explanations

phrases related to official statements or announcements

oai_token-act-pair · gpt-3.5-turbo

references to official endorsements or approvals from authoritative bodies

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

 Hacker

-0.63

 Blanc

-0.61

coh

-0.60

 Mankind

-0.57

anny

-0.57

 Strongh

-0.56

 Maxim

-0.56

Lad

-0.56

 Chung

-0.54

illes

-0.54

POSITIVE LOGITS

rawdownloadcloneembedreportprint

0.72

reb

0.72

ollah

0.71

oldown

0.70

uala

0.69

nces

0.67

DAQ

0.66

abeth

0.66

ulf

0.66

uca

0.65

Activations Density 0.464%

phrases related to official statements or announcements

references to official endorsements or approvals from authoritative bodies

No Comments

No Known Activations