INDEX

Explanations

ranking and classification terms related to popularity or success

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

Ranking something as number one or "#1"

np_acts-logits-general · gemini-2.0-flash

number one or top ranked

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-9b-it-res/layer_9/width_16k/average_l0_88

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.9.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

AnimationsModule

-0.48

Came

-0.41

 signora

-0.40

béco

-0.40

 Drafting

-0.39

Samples

-0.39

Dressing

-0.38

 seam

-0.38

duled

-0.37

lotz

-0.37

POSITIVE LOGITS

top

0.54

highest

0.52

 highest

0.52

 najbolj

0.51

 ranked

0.49

 tertinggi

0.48

 leading

0.48

Highest

0.47

 tops

0.47

best

0.47

Activations Density 0.351%

ranking and classification terms related to popularity or success

Ranking something as number one or "#1"

number one or top ranked

No Comments

No Known Activations