INDEX

Explanations

comparisons and contrasts present in 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

isf

-0.66

awar

-0.65

may

-0.64

haven

-0.63

afia

-0.62

auer

-0.62

cia

-0.61

Age

-0.61

WIND

-0.61

onen

-0.60

POSITIVE LOGITS

 necessarily

1.26

 merely

1.07

 simply

0.93

 relying

0.93

 bothering

0.90

 anything

0.84

 solely

0.81

 just

0.81

 letting

0.79

 mere

0.79

Activations Density 0.616%

comparisons and contrasts present in text

No Comments

No Known Activations