INDEX

Explanations

instances of bribery and related illegal activities

oai_token-act-pair · gpt-3.5-turbo

terms related to bribery and corruption

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

Adams

-0.75

Dub

-0.75

Trend

-0.74

noon

-0.74

LAN

-0.73

ISS

-0.73

ART

-0.73

nesota

-0.73

Saint

-0.73

brew

-0.73

POSITIVE LOGITS

 bribe

1.32

 bribes

1.20

 extortion

0.93

 bribery

0.88

 disgu

0.87

 bluff

0.82

 barg

0.81

 paid

0.80

 bait

0.80

 induce

0.78

Activations Density 0.008%

instances of bribery and related illegal activities

terms related to bribery and corruption

No Comments

No Known Activations