INDEX

Explanations

descriptions of receiving or lodging complaints or information

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_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

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

 mole

-0.61

 Strait

-0.60

apse

-0.60

iny

-0.60

olini

-0.58

 partying

-0.57

 fallacy

-0.57

 mortg

-0.57

mania

-0.55

hole

-0.54

POSITIVE LOGITS

 permission

0.90

 preferential

0.88

 notification

0.78

 assurances

0.77

DragonMagazine

0.76

 Citation

0.76

 royalties

0.75

 gifts

0.75

å¼

0.74

imental

0.74

Activations Density 0.503%

descriptions of receiving or lodging complaints or information

No Comments

No Known Activations