INDEX

Explanations

instances of relatedness or connections among different topics or articles

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

»Ĵ

-0.80

ainted

-0.76

UGE

-0.66

 clad

-0.65

Ĥİ

-0.65

ãĤ¡

-0.62

adra

-0.62

arest

-0.61

AFTA

-0.60

OF

-0.60

POSITIVE LOGITS

 Stories

1.09

 Coverage

0.89

 Content

0.85

 Links

0.83

 VIDEOS

0.82

 Picks

0.81

 ARTICLE

0.80

 Articles

0.80

 Video

0.79

 Advertisement

0.79

Activations Density 0.009%

instances of relatedness or connections among different topics or articles

No Comments

No Known Activations