INDEX

Explanations

conditional phrases indicating potential outcomes or requirements in situations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 36-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_36/width_16k/average_l0_120

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.36.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

fromnode

-0.59

'{@

-0.47

 CanadaChoose

-0.47

Personendaten

-0.47

 mérit

-0.46

Vidite

-0.46

 LoggerFactory

-0.44



-0.44

Contribute

-0.40

 насељу

-0.40

POSITIVE LOGITS

 apply

0.59

 replace

0.59

 treat

0.59

 reform

0.57

 substitute

0.54

 consider

0.53

 imposing

0.52

 replacing

0.52

 impose

0.52

 resourceCulture

0.51

Activations Density 2.031%

conditional phrases indicating potential outcomes or requirements in situations

No Comments

No Known Activations