INDEX

Explanations

references to leaked documents or information

oai_token-act-pair · gpt-3.5-turbo

instances of information related to leaks and whistleblowing activities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

oka

-0.74

 masc

-0.70

ciation

-0.70

 Manz

-0.68

wcsstore

-0.67

 Huss

-0.66

wagon

-0.66

NAD

-0.65

 Crow

-0.65

 Walls

-0.64

POSITIVE LOGITS

 hacked

1.31

 leaked

1.24

ileaks

1.23

 compromising

1.23

 trove

1.15

 secrets

1.10

 embarrassing

1.06

 confidential

1.03

 classified

1.02

DNC

1.02

Activations Density 0.220%

references to leaked documents or information

instances of information related to leaks and whistleblowing activities

No Comments

No Known Activations