INDEX

Explanations

phrases related to decision-making or planning

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_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

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

•How

manner1

No Comments

Negative Logits

 ........

-0.60

Joined

-0.59

odder

-0.59

iculture

-0.59

ola

-0.58

quist

-0.58

åĳ

-0.58

)]

-0.58

hari

-0.57

oubted

-0.57

POSITIVE LOGITS

soever

1.17

ells

0.91

 much

0.89

ls

0.88

ever

0.84

itzer

0.84

beit

0.80

 exactly

0.78

much

0.75

 badly

0.74

Activations Density 0.077%

phrases related to decision-making or planning

No Comments

No Known Activations