INDEX

Explanations

references to specific characters or titles from video games and related content

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

yrinth

-0.76

Wars

-0.73

ERAL

-0.68

idity

-0.64

inates

-0.64

eat

-0.63

inals

-0.62

lux

-0.62

ression

-0.61

vation

-0.59

POSITIVE LOGITS

sburg

0.82

yth

0.78

terson

0.77

ky

0.75

fee

0.75

burg

0.75

kins

0.74

hip

0.73

heed

0.72

 Glen

0.71

Activations Density 0.003%

references to specific characters or titles from video games and related content

No Comments

No Known Activations