INDEX

Explanations

instances of words related to selection or designation processes

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

chosen by or for

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.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

MatButtonModule

-0.45

بوابة

-0.40



-0.40

 procès

-0.38

paddingVertical

-0.38

Diweddarwch

-0.38

 comprando

-0.38

NewReader

-0.37

prodi

-0.37

 hours

-0.37

POSITIVE LOGITS

 chosen

0.78

 selected

0.75

 selection

0.73

chosen

0.73

 Selection

0.71

 Chosen

0.69

selection

0.66

selected

0.66

Chosen

0.62

选

0.62

Activations Density 0.013%

instances of words related to selection or designation processes

chosen by or for

No Comments

No Known Activations