INDEX

Explanations

words related to legal and law enforcement actions

oai_token-act-pair · gpt-3.5-turbo

mentions of actions involving social interactions or relationships

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 groundbreaking

-0.77

 DevOnline

-0.71

 International

-0.69

 unparalleled

-0.66

 Peak

-0.66

 Comprehensive

-0.64

 Bever

-0.64

 landmark

-0.64

 Featured

-0.63

 Pipeline

-0.63

POSITIVE LOGITS

they

1.19

 dont

1.11

 THEN

1.00

then

0.99

 THEY

0.98

morrow

0.93

 didnt

0.92

 obey

0.89

 doesnt

0.89

cause

0.89

Activations Density 0.677%

words related to legal and law enforcement actions

mentions of actions involving social interactions or relationships

No Comments

No Known Activations