INDEX

Explanations

phrases related to personal preferences and decisions

oai_token-act-pair · gpt-3.5-turbo 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

¬¼

-0.72

iphate

-0.66

mx

-0.65

xes

-0.64

iciency

-0.63

ulz

-0.62

emp

-0.61

emn

-0.61

uble

-0.61

encies

-0.60

POSITIVE LOGITS

 namely

1.68

viz

1.19

 albeit

1.01

 including

0.94

but

0.86

 plus

0.85

 minus

0.82

 resulting

0.81

but

0.81

 except

0.78

Activations Density 0.350%

phrases related to personal preferences and decisions

No Comments

No Known Activations