INDEX

Explanations

words related to statistics and rates

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_fs1536-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

1,536

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

sonian

-0.99

olog

-0.71

ĪĴ

-0.70

vell

-0.70

iverse

-0.70

ovie

-0.70

¬¼

-0.67

lear

-0.67

otom

-0.67

JECT

-0.65

POSITIVE LOGITS

 ratio

1.31

 ratios

1.27

 Ratio

1.05

 fluctuations

1.03

 peaked

1.02

 estimation

0.99

 coefficient

0.98

 rate

0.97

 measurement

0.96

 percentage

0.96

Activations Density 2.376%

words related to statistics and rates

No Comments

No Known Activations