INDEX

Explanations

feedback and interaction phrases related to user engagement

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

bably

-0.68

bryce

-0.62

otom

-0.60

aturally

-0.58

elong

-0.58

eating

-0.55

taboola

-0.54

Murray

-0.54

Ö¼

-0.54

inventoryQuantity

-0.54

POSITIVE LOGITS

 Credits

0.94

 Documentation

0.91

 Contents

0.88

FAQ

0.86

 Tutorial

0.85

 Requirements

0.85

 Installation

0.84

FAQ

0.83

 Definitions

0.83

 Authors

0.82

Activations Density 0.270%

feedback and interaction phrases related to user engagement

No Comments

No Known Activations