INDEX

Explanations

phrases that indicate the source of information or statements

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

introduces origin or context

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-131k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

 Италијани

-0.55

NOPQRST

-0.50

__*/

-0.50

 iprot

-0.49

ValueStyle

-0.49

MessageOf

-0.49

 gynhyrchwyd

-0.47

pogon

-0.47

irited

-0.46

bastien

-0.45

POSITIVE LOGITS

From

0.77

 From

0.75

Dari

0.47

FROM

0.46

 Desde

0.45

Från

0.44

from

0.43

從

0.43

 from

0.43

 Từ

0.42

Activations Density 0.016%

phrases that indicate the source of information or statements

introduces origin or context

No Comments

No Known Activations