INDEX

Explanations

phrases indicating a point in time or perspective

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

 twin

-0.68

 twins

-0.63

amaz

-0.61

harm

-0.61

 Cities

-0.60

ours

-0.59

lez

-0.58

discrimination

-0.56

 warranties

-0.56

 subst

-0.56

POSITIVE LOGITS

cture

0.86

 onwards

0.83

 onward

0.77

ebin

0.70

abouts

0.70

endment

0.69

ozo

0.63

ajor

0.63

gue

0.62

EStreamFrame

0.61

Activations Density 0.035%

phrases indicating a point in time or perspective

No Comments

No Known Activations