INDEX

Explanations

references to refactoring and refining code

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_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

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

istically

-0.78

ctica

-0.75

ties

-0.74

ahime

-0.74

 Tsarnaev

-0.71

istic

-0.71

amaru

-0.71

alian

-0.70

chin

-0.69

owski

-0.68

POSITIVE LOGITS

eree

1.08

lection

0.99

ractive

0.92

riger

0.90

erential

0.87

lections

0.85

eren

0.81

erer

0.80

raction

0.79

ractor

0.78

Activations Density 1.268%

references to refactoring and refining code

No Comments

No Known Activations