INDEX

Explanations

mentions related to legal issues and corruption cases involving high-profile individuals

oai_token-act-pair · gpt-3.5-turbo

topics related to high-profile corruption cases

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

áµ

-0.95

 Sioux

-0.79

colored

-0.78

 Zombies

-0.77

 Prairie

-0.76

Favorite

-0.76

Kyle

-0.76

¥µ

-0.76

aukee

-0.74

Texas

-0.74

POSITIVE LOGITS

 scandals

1.60

 bribery

1.59

 bribes

1.58

 scandal

1.53

 corruption

1.52

 bribe

1.44

 allegations

1.39

corruption

1.38

 whistleblower

1.37

 Corruption

1.37

Activations Density 0.810%

mentions related to legal issues and corruption cases involving high-profile individuals

topics related to high-profile corruption cases

No Comments

No Known Activations