INDEX

Explanations

phrases related to deception and falsification

oai_token-act-pair · gpt-3.5-turbo

keywords related to deception and fraud

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

odder

-0.84

itar

-0.80

âĶĢâĶĢâĶĢâĶĢâĶĢâĶĢâĶĢâĶĢ

-0.80

%:

-0.79

âĶģ

-0.76

gdala

-0.73

ãĥ¼ãĤ¯

-0.69

gard

-0.69

intosh

-0.69

edited

-0.68

POSITIVE LOGITS

 oath

0.92

 payments

0.86

 paperwork

0.85

0.83

 registrations

0.79

 payment

0.78

 vaccinations

0.77

 promises

0.77

 documents

0.77

 exams

0.76

Activations Density 0.243%

phrases related to deception and falsification

keywords related to deception and fraud

No Comments

No Known Activations