INDEX

Explanations

references to badges and their significance

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

noon

-0.82

theless

-0.80

pter

-0.74

apest

-0.73

ydia

-0.73

moving

-0.72

frey

-0.72

reb

-0.72

VEN

-0.71

birth

-0.69

POSITIVE LOGITS

 badges

1.15

 badge

1.11

 insign

1.09

 Badge

0.92

 holder

0.85

 worn

0.84

 bearer

0.83

 buckle

0.81

 wearer

0.79

 collar

0.79

Activations Density 0.008%

references to badges and their significance

No Comments

No Known Activations