INDEX

Explanations

mentions of technical and scientific terms or concepts

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_fs1536-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

1,536

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

ratulations

-0.75

iversary

-0.75

pheus

-0.74

 Congratulations

-0.74

Congratulations

-0.73

vana

-0.73

ACP

-0.72

Happy

-0.72

redible

-0.72

DragonMagazine

-0.72

POSITIVE LOGITS

 inability

1.47

 inadequ

1.32

 inefficient

1.28

 inadequate

1.28

 Unable

1.26

 inconsistent

1.26

 unable

1.25

 inexper

1.25

 worsened

1.24

 deteriorated

1.23

Activations Density 5.855%

mentions of technical and scientific terms or concepts

No Comments

No Known Activations