INDEX

Explanations

proper nouns related to movies and entertainment

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_fs1536-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

1,536

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

undai

-0.77

 troubles

-0.73

izons

-0.71

 tremend

-0.71

odox

-0.71

leased

-0.69

oud

-0.69

etooth

-0.68

 trouble

-0.68

bered

-0.67

POSITIVE LOGITS

 Provided

1.07

 https

0.89

 http

0.86

 Become

0.80

 Anyone

0.76

YES

0.75

 Cosponsors

0.73

 Unknown

0.73

Bye

0.73

 Countries

0.72

Activations Density 0.605%

proper nouns related to movies and entertainment

No Comments

No Known Activations