INDEX

Explanations

references to achievements and victories

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

win, winner, winning

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

 arşivlendi

-0.43

ImageContext

-0.43

 referrerpolicy

-0.40

 limousine

-0.37

 dessen

-0.37

tdessen

-0.36

NewUrlParser

-0.35

 forState

-0.35

disc

-0.34

 useRef

-0.33

POSITIVE LOGITS

win

1.19

 winning

1.13

 ganar

1.13

 winner

1.11

 gain

1.10

winning

1.07

Winner

1.06

 Winning

1.06

Winning

1.05

 gewinnen

1.02

Activations Density 0.014%

references to achievements and victories

win, winner, winning

No Comments

No Known Activations