INDEX

Explanations

impolite or offensive language

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

uate

-0.82

 unpre

-0.68

 Prelude

-0.67

 inhibition

-0.64

 headache

-0.59

 gratification

-0.59

 duplication

-0.59

 conscientious

-0.59

psy

-0.58

 susceptibility

-0.57

POSITIVE LOGITS

ITNESS

1.28

atts

1.25

orthy

1.20

idespread

1.19

atson

1.18

reck

1.17

affle

1.16

restling

1.16

elcome

1.16

edge

1.15

Activations Density 0.326%

impolite or offensive language

No Comments

No Known Activations