INDEX

Explanations

phrases related to cooperation with law enforcement or authorities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs49152-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 horm

-0.72

pring

-0.72

ãĤ¦ãĤ¹

-0.66

ortality

-0.65

ãĥĩãĤ£

-0.62

glass

-0.62

 teasp

-0.61

nu

-0.61

stall

-0.61

mill

-0.60

POSITIVE LOGITS

 prosecution

0.93

 efforts

0.89

 investigations

0.87

 authorities

0.83

 extradition

0.82

 investigation

0.81

 investigators

0.80

 prosecutors

0.78

 disclosure

0.78

 inquiries

0.77

Activations Density 0.050%

phrases related to cooperation with law enforcement or authorities

No Comments

No Known Activations