INDEX

Explanations

phrases that describe names or titles given to entities or concepts

oai_token-act-pair · gpt-4o-mini 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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ython

-0.77

 Flavoring

-0.74

igsaw

-0.71

onential

-0.69

omas

-0.66

ike

-0.63

restling

-0.63

acca

-0.63

 disarm

-0.62

anqu

-0.62

POSITIVE LOGITS

 wont

0.80

ÑĮ

0.71

iHUD

0.69

)[

0.67

 writ

0.66

displayText

0.65

pects

0.65

 fare

0.64

 dictates

0.63

pires

0.62

Activations Density 0.108%

phrases that describe names or titles given to entities or concepts

No Comments

No Known Activations