INDEX

Explanations

location-related phrases, names of places, and contexts involving community services and businesses

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

wordpress

-0.72

ĸļ

-0.72

ulet

-0.70

ŀ

-0.70

 Emin

-0.70

illac

-0.70

 RELE

-0.68

Ùħ

-0.67

grave

-0.67

pter

-0.66

POSITIVE LOGITS

strument

1.10

vention

0.89

iti

0.88

jured

0.85

verted

0.81

hib

0.81

hibit

0.80

hibited

0.80

clusive

0.80

cluded

0.79

Activations Density 0.020%

location-related phrases, names of places, and contexts involving community services and businesses

No Comments

No Known Activations