INDEX

Explanations

phrases related to capability and limitations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 7-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_7/width_16k/average_l0_69

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.7.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

igshid

-0.69

uzzi

-0.54

iNdEx

-0.53

 tärke

-0.49

endedor

-0.48

 uska

-0.45

 semper

-0.44

fusc

-0.44

Unmarshaller

-0.42

enterOuterAlt

-0.42

POSITIVE LOGITS

 impossible

1.26

 achievable

1.15

 Impossible

1.15

impossible

1.14

Impossible

1.14

 doable

1.09

 impossibility

1.08

 attainable

1.02

 imposible

1.00

featureID

0.88

Activations Density 0.364%

phrases related to capability and limitations

No Comments

No Known Activations