INDEX

Explanations

mentions of proper nouns or names in particular contexts

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

among

-0.71

lege

-0.71

stellar

-0.65

craft

-0.64

rand

-0.64

ordinate

-0.63

ftime

-0.63

onse

-0.63

 respecting

-0.62

ucl

-0.62

POSITIVE LOGITS

 same

1.07

 proverbial

1.05

 brakes

0.97

 hardest

0.95

 pavement

0.89

 ground

0.88

 shores

0.86

 latter

0.85

 heavens

0.85

 brink

0.84

Activations Density 0.097%

mentions of proper nouns or names in particular contexts

No Comments

No Known Activations