INDEX

Explanations

references to significant or impactful events, actions, or observations

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

agents

-0.78

Own

-0.72

 Mysteries

-0.71

Ord

-0.71

achu

-0.67

 Jagu

-0.65

Contents

-0.65

ï

-0.65

DIV

-0.65

agree

-0.64

POSITIVE LOGITS

 handful

1.23

 consequ

1.20

 bunch

1.11

 slew

1.08

few

1.06

lot

1.04

 penchant

1.04

 plethora

1.04

cknowled

1.03

 corresponding

1.03

Activations Density 0.114%

references to significant or impactful events, actions, or observations

No Comments

No Known Activations