INDEX

Explanations

references to the United States and immigration status

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.8.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

panel

-0.72

Panel

-0.66

utm

-0.66

sbm

-0.63

catentry

-0.62

 Mats

-0.62

Nap

-0.61

Wem

-0.61

WHO

-0.60

Corp

-0.60

POSITIVE LOGITS

 illegally

1.43

 unlawfully

1.19

 undet

1.06

 legally

0.98

 unaccompanied

0.94

 safely

0.88

 permanently

0.86

 lawfully

0.86

 voluntarily

0.82

 bloodstream

0.82

Activations Density 0.173%

references to the United States and immigration status

No Comments

No Known Activations