INDEX

Explanations

technical terms related to software development and programming

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_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

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

 beforehand

-0.73

 prior

-0.63

eret

-0.63

 during

-0.62

prov

-0.62

 intimid

-0.58

 earlier

-0.58

IGN

-0.57

 Kislyak

-0.57

while

-0.56

POSITIVE LOGITS

 finally

1.06

 again

0.92

 reckoning

0.91

 realize

0.87

 anew

0.86

 relent

0.82

 realise

0.81

 recons

0.81

nir

0.80

again

0.79

Activations Density 2.555%

technical terms related to software development and programming

No Comments

No Known Activations