INDEX

Explanations

mentions of locations, authorities, and legal terms

oai_token-act-pair · gpt-3.5-turbo

titles and roles of individuals, particularly in a law enforcement or official context

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

colored

-0.90

defense

-0.89

 favors

-0.88

 honored

-0.87

avored

-0.86

 rumors

-0.86

 honor

-0.85

 honoring

-0.84

 unfavorable

-0.84

 tumors

-0.84

POSITIVE LOGITS

 Â£

1.61

 Edinburgh

1.39

 Yorkshire

1.39

NHS

1.39

 Scotland

1.37

 Cardiff

1.37

Â£

1.35

Scotland

1.34

 Ukip

1.32

 Labour

1.31

Activations Density 3.471%

mentions of locations, authorities, and legal terms

titles and roles of individuals, particularly in a law enforcement or official context

No Comments

No Known Activations