INDEX

Explanations

phrases related to physical actions and gestures

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

 Lowell

-0.66

 Atmosp

-0.65

 adversely

-0.62

ļéĨĴ

-0.62

 mmol

-0.61

 Belle

-0.61

uary

-0.61

 Everest

-0.60

 Salvation

-0.60

Ń·

-0.59

POSITIVE LOGITS

picked

1.26

crafted

1.17

held

1.16

washing

1.11

selected

1.11

made

1.09

built

1.07

painted

1.07

drawn

1.07

eye

1.07

Activations Density 0.010%

phrases related to physical actions and gestures

No Comments

No Known Activations