INDEX

Explanations

copyright symbols and associated attributes or information

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

New Auto-Interp

Configuration

ctigges/pythia-70m-deduped__res-sm_processed/1-res-sm

Prompts (Dashboard)

32,768 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

32,768

Data Type

torch.float32

Hook Name

blocks.1.hook_resid_post

Hook Layer

Architecture

standard

Context Size

128

Dataset

EleutherAI/the_pile_deduplicated

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 easiest

-1.58

 used

-1.53

 averaged

-1.52

 pitched

-1.50

where

-1.48

liest

-1.43

 sustaining

-1.39

 capital

-1.38

 averaging

-1.38

 compared

-1.38

POSITIVE LOGITS

 Conclusions

1.69

 Redistributions

1.65

 Conclusion

1.64

Ĭ

1.63

us

1.62

suppl

1.57

ORANDUM

1.56

biamo

1.55

 NEGLIGENCE

1.53

 Contents

1.52

Activations Density 0.421%

copyright symbols and associated attributes or information

No Comments

No Known Activations