INDEX

Explanations

numerical figures and quantities related to various topics

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

rium

-0.73

liness

-0.63

aban

-0.62

timer

-0.59

ieu

-0.59

aeus

-0.58

pper

-0.58

estine

-0.57

amaru

-0.57

flush

-0.57

POSITIVE LOGITS

 consecutive

1.14

 percent

1.06

 million

1.03

1.01

 thousand

0.90

 different

0.88

 acres

0.88

 pieces

0.87

 instances

0.86

Activations Density 0.957%

numerical figures and quantities related to various topics

No Comments

No Known Activations