INDEX

Explanations

phrases related to significant accomplishments or characteristics

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

 oneself

-0.76

ÃŁ

-0.69

orks

-0.62

 yourselves

-0.61

 Federation

-0.60

 Yose

-0.60

 Brunswick

-0.59

annon

-0.58

 Saud

-0.58

Spawn

-0.58

POSITIVE LOGITS

mates

0.85

mate

0.84

 counterparts

0.83

fulness

0.80

lessness

0.79

 abroad

0.79

 foray

0.78

ings

0.76

 exploits

0.76

 spree

0.76

Activations Density 1.615%

phrases related to significant accomplishments or characteristics

No Comments

No Known Activations