INDEX

Explanations

words related to the removal and implications of lead in various 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_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

bags

-0.79

xual

-0.79

ership

-0.76

boxes

-0.72

ery

-0.68

ers

-0.65

STEM

-0.64

pot

-0.63

Ð²

-0.63

wit

-0.62

POSITIVE LOGITS

aining

0.93

ainer

0.92

ittance

0.91

bered

0.86

icably

0.85

ained

0.85

raction

0.85

oving

0.84

illard

0.84

inant

0.81

Activations Density 0.804%

words related to the removal and implications of lead in various contexts

No Comments

No Known Activations