INDEX

Explanations

language related to support and assistance like "Privacy Policy" and "Sign up."

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

abal

-0.67

agi

-0.63

ebted

-0.63

iaries

-0.58

ovember

-0.58

vill

-0.57

 Himself

-0.54

%:

-0.53

uable

-0.53

 Reincarnated

-0.52

POSITIVE LOGITS

 prud

0.65

{*

0.59

="/

0.57

orpor

0.56

lax

0.56

ĨĴ

0.55

 rash

0.54

ennis

0.52

 constitution

0.51

 hesitated

0.51

Activations Density 0.047%

language related to support and assistance like "Privacy Policy" and "Sign up."

No Comments

No Known Activations