INDEX

Explanations

words related to significant numerical counts and measurements

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

 Diary

-0.75

 Karl

-0.74

 Hearts

-0.71

 Connor

-0.70

twe

-0.70

 chick

-0.69

 Beth

-0.69

McD

-0.68

Jew

-0.68

 ensu

-0.67

POSITIVE LOGITS

ur

1.72

ura

1.38

ure

1.37

uran

1.36

urs

1.35

ury

1.35

uron

1.33

UR

1.33

ural

1.32

uria

1.30

Activations Density 0.064%

words related to significant numerical counts and measurements

No Comments

No Known Activations