INDEX

Explanations

references to the word "there" and its variations

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

tein

-0.79

NAS

-0.79

STON

-0.73

dor

-0.71

vier

-0.69

wordpress

-0.67

Unix

-0.67

tnc

-0.67

enegger

-0.66

Disk

-0.66

POSITIVE LOGITS

abouts

1.07

upon

0.76

 afterwards

0.73

 throughout

0.72

 quizz

0.69

 thereafter

0.68

 else

0.68

 glimps

0.65

 afterward

0.65

 whilst

0.64

Activations Density 0.008%

references to the word "there" and its variations

No Comments

No Known Activations

references to the word "there" and its variations

No Comments

No Known Activations