INDEX

Explanations

descriptions of physical appearances and personalities, with a focus on attributes like demeanor, confidence, and demeanor

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

SEA

-0.89

URA

-0.79

chwitz

-0.76

isible

-0.71

undo

-0.71

OTOS

-0.70

socket

-0.70

RF

-0.69

Ô

-0.69

significant

-0.69

POSITIVE LOGITS

ness

1.17

 grin

1.02

nesses

0.98

 tendencies

0.97

 tones

0.94

ly

0.92

 demeanor

0.92

 humour

0.91

 glances

0.88

 beginnings

0.85

Activations Density 1.124%

descriptions of physical appearances and personalities, with a focus on attributes like demeanor, confidence, and demeanor

No Comments

No Known Activations