INDEX

Explanations

discussions about pronunciation and the related complexities of linguistic forms

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs49152-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

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

ansas

-0.79

iday

-0.76

 Frie

-0.70

 Volunte

-0.68

+)

-0.66

ashtra

-0.66

 comprom

-0.65

ionage

-0.65

waukee

-0.64

 Airl

-0.64

POSITIVE LOGITS

unci

1.20

ciation

1.02

 pronunciation

0.94

 pronounce

0.92

TY

0.89

ouncing

0.89

 pronoun

0.88

phas

0.80

casters

0.78

ters

0.77

Activations Density 0.005%

discussions about pronunciation and the related complexities of linguistic forms

No Comments

No Known Activations