INDEX

Explanations

sequences of stars and decorative symbols

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

 artificially

-0.61

wagen

-0.59

bda

-0.59

pard

-0.59

bledon

-0.59

 levers

-0.58

 slic

-0.57

 bast

-0.57

 velvet

-0.57

 brute

-0.56

POSITIVE LOGITS

--------------------------------------------------------

0.81

 CONTIN

0.78

 ********************************

0.77

CHAPTER

0.76

 Chapters

0.74

ERROR

0.73

0.71

THE

0.70

======

0.70

 WATCH

0.69

Activations Density 0.006%

sequences of stars and decorative symbols

No Comments

No Known Activations