INDEX

Explanations

occurrences of the word "the"

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

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

<?

-0.79

âĺ

-0.75

among

-0.75

heit

-0.74

lly

-0.72

HO

-0.72

FOR

-0.71

ãĥ¯

-0.66

zbek

-0.66

eous

-0.66

POSITIVE LOGITS

 expiration

1.11

 onset

1.07

 initial

1.06

 arrival

1.05

 completion

1.02

 breakup

0.99

 collapse

0.98

 fact

0.96

 introduction

0.96

 conclusion

0.96

Activations Density 0.103%

occurrences of the word "the"

No Comments

No Known Activations