INDEX

Explanations

terms related to vandalism and graffiti

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs49152-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

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

vana

-0.85

dinand

-0.81

ramer

-0.78

amacare

-0.77

croft

-0.77

hart

-0.75

xus

-0.74

lopp

-0.73

travel

-0.72

berman

-0.72

POSITIVE LOGITS

 vandalism

1.14

 spree

1.08

 graffiti

0.97

 vandal

0.88

 rampage

0.87

 Alley

0.76

andals

0.74

 perpetrated

0.72

 inflicted

0.72

 leve

0.70

Activations Density 0.015%

terms related to vandalism and graffiti

No Comments

No Known Activations