INDEX

Explanations

phrases related to reviewing or evaluating products and services

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_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

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

zbollah

-0.21

 Âł

-0.21

withstanding

-0.21

 Originally

-0.21

 Belfast

-0.21

 Chin

-0.20

NK

-0.20

è£ħ

-0.20

idas

-0.20

gged

-0.20

POSITIVE LOGITS

Malley

0.19

Flow

0.19

ONSORED

0.19

alone

0.19

manac

0.18

 afterward

0.18

 outwe

0.18

powered

0.18

 hers

0.18

 appliances

0.18

Activations Density 13.094%

phrases related to reviewing or evaluating products and services

No Comments

No Known Activations