INDEX

Explanations

phrases referring to parts or portions of something

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

Reviewer

-0.67

quartered

-0.63

 Blues

-0.62

 Commando

-0.59

MSN

-0.57

asma

-0.57

axies

-0.53

 elig

-0.53

iform

-0.52

top

-0.52

POSITIVE LOGITS

dozen

0.80

 century

0.69

decade

0.68

çĶ°

0.66

 YEAR

0.65

 millennia

0.64

 dozen

0.64

imester

0.63

lot

0.61

maxwell

0.61

Activations Density 0.029%

phrases referring to parts or portions of something

No Comments

No Known Activations