INDEX

Explanations

companies, organizations, and programs

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

ĨĴ

-0.71

 cellul

-0.65

llah

-0.62

WD

-0.60

Ö¼

-0.59

catentry

-0.58

 stewards

-0.55

 salsa

-0.54

 thous

-0.54

lur

-0.53

POSITIVE LOGITS

emort

0.86

heed

0.77

oyd

0.74

gow

0.74

utan

0.72

ocene

0.72

ativity

0.71

ansk

0.71

monds

0.70

ograp

0.69

Activations Density 3.266%

companies, organizations, and programs

No Comments

No Known Activations