INDEX

Explanations

decade references in the text

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

edIn

-0.79

ournals

-0.79

 snipp

-0.69

acts

-0.68

Name

-0.65

Tickets

-0.64

stro

-0.64

azor

-0.63

Maker

-0.63

Weak

-0.63

POSITIVE LOGITS

 onwards

1.23

 onward

1.15

era

1.12

Era

1.02

 nostalg

0.93

BCE

0.92

ixties

0.89

-'

0.83

hey

0.81

 Revival

0.77

Activations Density 0.562%

decade references in the text

No Comments

No Known Activations