INDEX

Explanations

scientific or technical terms commonly found in research articles

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

"]=>

-0.74

SELECT

-0.70

Firstly

-0.65

meet

-0.64

obin

-0.63

lly

-0.61

blocks

-0.60

oster

-0.60

UID

-0.60

issue

-0.60

POSITIVE LOGITS

Jr

1.13

maxwell

0.83

tein

0.74

formerly

0.72

Jr

0.71

aeus

0.70

QC

0.70

 retired

0.69

Cha

0.69

Sr

0.68

Activations Density 0.146%

scientific or technical terms commonly found in research articles

No Comments

No Known Activations