INDEX

Explanations

words related to specific series titles, specifically "Cloak & Dagger"

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

tumblr

-0.76

xual

-0.72

icist

-0.69

drm

-0.68

urgical

-0.66

hov

-0.65

ufact

-0.64

anie

-0.63

azeera

-0.63

soDeliveryDate

-0.62

POSITIVE LOGITS

ãĤĵ

0.69

Emer

0.67

ridge

0.63

Upon

0.61

 bounds

0.61

 Berry

0.60

Lay

0.59

Forest

0.59

Quotes

0.58

ãģŁ

0.58

Activations Density 3.400%

words related to specific series titles, specifically "Cloak & Dagger"

No Comments

No Known Activations