INDEX

Explanations

phrases indicating the prevalence or dissent of opinions among a group of people

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

 ................

-0.74

cule

-0.68

 Reincarn

-0.65

ĸļ

-0.63

inders

-0.63

]+

-0.62

almost

-0.62

Mub

-0.62

 Hover

-0.61

Hmm

-0.61

POSITIVE LOGITS

 necessarily

0.74

 appe

0.70

 subscrib

0.69

asio

0.67

 interested

0.67

 agree

0.67

 effected

0.67

 affected

0.66

 agrees

0.65

 audience

0.65

Activations Density 0.013%

phrases indicating the prevalence or dissent of opinions among a group of people

No Comments

No Known Activations