INDEX

Explanations

phrases indicating presence or status

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

is/are/was/were followed by adjective or noun

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

tap

-0.32

 giây

-0.32

 Medium

-0.29

 scheme

-0.28

Legături

-0.28

它

-0.28

dtp

-0.26

 Ähn

-0.26

Tap

-0.26

 τη

-0.25

POSITIVE LOGITS

帖最后由

0.64

 Menſchen

0.63

 deſſen

0.63

 dieſen

0.63

rungsseite

0.63

complexType

0.62

 ſua

0.62

ſſung

0.62

 ſeinen

0.61

IntoConstraints

0.61

Activations Density 0.054%

phrases indicating presence or status

is/are/was/were followed by adjective or noun

No Comments

No Known Activations