INDEX

Explanations

proper nouns related to characters or entities

oai_token-act-pair · gpt-3.5-turbo

references to a specific character or entity, possibly "Gro", in various 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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

CPC

-0.74

 null

-0.65

 writ

-0.62

CTR

-0.59

 Gravity

-0.58

 descendant

-0.58

 NULL

-0.57

 Fully

-0.57

AAP

-0.57

 relation

-0.56

POSITIVE LOGITS

tesque

1.43

cery

1.30

ove

1.16

lean

1.13

ovy

1.07

veland

1.06

eger

1.02

gan

1.02

bie

1.01

ogle

1.00

Activations Density 0.025%

proper nouns related to characters or entities

references to a specific character or entity, possibly "Gro", in various contexts

No Comments

No Known Activations