INDEX

Explanations

words associated with burdens and emotional weight

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

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

 Scotia

-0.74

terday

-0.65

agents

-0.63

 interstate

-0.61

 Prosecut

-0.61

Downloadha

-0.60

HRC

-0.57

utherford

-0.56

CTR

-0.56

POSITIVE LOGITS

ened

2.08

ening

1.92

eners

1.61

enment

1.33

ens

1.33

ener

1.29

en

1.24

nesses

1.02

eling

0.97

eled

0.94

Activations Density 0.040%

words associated with burdens and emotional weight

No Comments

No Known Activations