INDEX

Explanations

phrases related to actions or events occurring in a sequential order

oai_token-act-pair · gpt-3.5-turbo

the use of the word "then" in narrative contexts

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.9.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

 Presence

-0.69

 belief

-0.68

 Doctrine

-0.66

 competence

-0.61

 behavi

-0.61

BAT

-0.60

 Transactions

-0.59

 norms

-0.59

orum

-0.58

 exception

-0.58

POSITIVE LOGITS

 proceeded

1.32

 proceed

0.88

 proceeds

0.87

 transitioned

0.86

 promptly

0.85

forth

0.84

 turned

0.80

DragonMagazine

0.77

 marched

0.77

 embarked

0.77

Activations Density 0.052%

phrases related to actions or events occurring in a sequential order

the use of the word "then" in narrative contexts

No Comments

No Known Activations