INDEX

Explanations

references to technical and possibly security-related terms or concepts

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

ensibly

-0.58

 contingency

-0.56

omics

-0.54

 credential

-0.51

etta

-0.50

 facilit

-0.50

 migration

-0.49

 prud

-0.48

 nonprofit

-0.48

 legislative

-0.48

POSITIVE LOGITS

________________

0.78

--------------------------------------------------------

0.76

[/

0.76

http

0.76

----------------------------------------------------------------

0.74

 ********************************

0.73

================================================================

0.72

________________________________________________________________

0.71

--------------------------------

0.69

 http

0.69

Activations Density 1.010%

references to technical and possibly security-related terms or concepts

No Comments

No Known Activations