INDEX

Explanations

information related to criminal activities, investigations, and legal proceedings

oai_token-act-pair · gpt-3.5-turbo 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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

caliber

-0.87

colored

-0.86

defense

-0.86

avored

-0.86

imize

-0.85

 favors

-0.85

 someday

-0.84

 demeanor

-0.83

Defense

-0.81

 gotten

-0.81

POSITIVE LOGITS

 Â£

1.54

 Labour

1.49

 recognise

1.48

 Ukip

1.45

 UKIP

1.41

 Scotland

1.40

 realise

1.40

 apologised

1.36

Labour

1.34

 centre

1.34

Activations Density 2.254%

information related to criminal activities, investigations, and legal proceedings

No Comments

No Known Activations