INDEX

Explanations

words related to legal actions and processes, such as "dropped," "canceled," "scrapped," and "withdrawn."

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

Fit

-0.74

rouse

-0.72

idth

-0.71

rics

-0.61

Tang

-0.61

PRES

-0.59

 Rhythm

-0.58

Strength

-0.58

 Luke

-0.58

ses

-0.57

POSITIVE LOGITS

 altogether

1.14

 entirely

1.00

 abruptly

0.97

 indefinitely

0.95

uled

0.86

 unanimously

0.86

by

0.85

 owing

0.85

 administr

0.85

 amid

0.83

Activations Density 0.147%

words related to legal actions and processes, such as "dropped," "canceled," "scrapped," and "withdrawn."

No Comments

No Known Activations