INDEX

Explanations

information related to weight loss studies and research

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs768-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_768

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

768

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

Contents

-0.77

evidence

-0.76

anism

-0.75

Edit

-0.75

grounds

-0.73

agree

-0.71

Events

-0.70

iments

-0.69

Examples

-0.69

words

-0.69

POSITIVE LOGITS

lot

1.32

 bunch

1.29

 handful

1.24

 plethora

1.23

 multitude

1.16

 couple

1.12

 slew

1.08

few

1.08

 whopping

1.07

 dozen

1.06

Activations Density 2.440%

information related to weight loss studies and research

No Comments

No Known Activations