INDEX

Explanations

references to shadows and their symbolic implications

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

XY

-0.71

ongs

-0.63

 preferably

-0.63

nice

-0.63

atography

-0.62

Methods

-0.62

Fan

-0.60

 fixme

-0.60

Favorite

-0.60

uggest

-0.59

POSITIVE LOGITS

 cloud

0.76

 looming

0.76

 looms

0.71

belt

0.70

leader

0.68

oubt

0.68

eless

0.67

 shroud

0.67

 canopy

0.66

 gloom

0.66

Activations Density 0.046%

references to shadows and their symbolic implications

No Comments

No Known Activations