INDEX

Explanations

number sequences within text

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_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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

itably

-0.75

thirds

-0.73

 gain

-0.69

 grounds

-0.66

 warranties

-0.64

etimes

-0.64

ãĤ¶

-0.64

 pasture

-0.62

 Commonwealth

-0.62

 statutes

-0.62

POSITIVE LOGITS

DragonMagazine

0.86

CB

0.81

Reviewer

0.79

Quality

0.76

 Recommend

0.70

 Rated

0.69

 Quote

0.69

0.68

NK

0.68

0.67

Activations Density 0.029%

number sequences within text

No Comments

No Known Activations