INDEX

Explanations

locations or specific buildings, often related to government or organizations

oai_token-act-pair · gpt-3.5-turbo

references to law enforcement and governmental institutions

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

ationally

-0.69

idae

-0.67

ealous

-0.65

married

-0.64

intent

-0.64

Els

-0.63

lishes

-0.62

pse

-0.61

independent

-0.60

vae

-0.60

POSITIVE LOGITS

 headquarters

1.63

 offices

1.58

 office

1.39

 Headquarters

1.34

HQ

1.30

 hangar

1.29

 building

1.29

 mansion

1.28

 courthouse

1.27

 residence

1.25

Activations Density 0.254%

locations or specific buildings, often related to government or organizations

references to law enforcement and governmental institutions

No Comments

No Known Activations