INDEX

Explanations

references to identification numbers or IDs

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 sidewalks

-0.69

ĵĺ

-0.69

andise

-0.68

 antioxid

-0.68

SPONSORED

-0.66

 pollen

-0.66

 untreated

-0.66

bilt

-0.66

 shotguns

-0.65

orial

-0.65

POSITIVE LOGITS

ividual

1.01

aho

1.00

entity

0.99

irection

0.95

iots

0.95

iom

0.94

irect

0.91

gin

0.88

ictionary

0.86

0.83

Activations Density 0.013%

references to identification numbers or IDs

No Comments

No Known Activations