INDEX

Explanations

descriptions related to scientific discoveries and geographical features

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

ippery

-0.68

IPP

-0.67

Pak

-0.66

hov

-0.66

ritical

-0.65

icity

-0.64

asking

-0.64

ovy

-0.64

henko

-0.64

nown

-0.63

POSITIVE LOGITS

 therein

0.76

 herein

0.74

 today

0.72

GD

0.70

 humankind

0.68

 mankind

0.67

 Mankind

0.66

 Rubin

0.66

Alc

0.65

viz

0.65

Activations Density 2.297%

descriptions related to scientific discoveries and geographical features

No Comments

No Known Activations