INDEX

Explanations

references to cartoons and animated media

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

 opio

-0.70

acia

-0.66

forces

-0.66

 sclerosis

-0.64

govern

-0.64

vae

-0.63

pta

-0.62

CI

-0.60

kept

-0.59

-0.58

POSITIVE LOGITS

ishly

1.19

 cartoons

1.03

ish

1.00

ists

1.00

ist

0.97

oons

0.90

 cartoon

0.90

 caric

0.89

ical

0.89

istically

0.87

Activations Density 0.015%

references to cartoons and animated media

No Comments

No Known Activations