INDEX

Explanations

actions related to physical interactions and movements

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

lihood

-0.73

else

-0.63

 incentiv

-0.62

MAT

-0.61

 hampered

-0.60

yond

-0.60

NG

-0.60

 complement

-0.59

etheless

-0.58

ification

-0.57

POSITIVE LOGITS

 herself

1.02

 himself

0.82

 candles

0.78

rily

0.74

lasses

0.73

 knives

0.71

her

0.69

 oneself

0.68

 copies

0.68

brush

0.68

Activations Density 0.444%

actions related to physical interactions and movements

No Comments

No Known Activations