INDEX

Explanations

connections related to actions, intentions, and effectiveness within various contexts

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

best

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 31-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_31/width_131k/average_l0_109

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.31.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

 betweenstory

-0.51

 mergeFrom

-0.47

yyb

-0.47

blame

-0.47

RenderAtEndOf

-0.42

 Roskov

-0.40

smoking

-0.40

 Exposure

-0.40

inghouse

-0.39

omock

-0.39

POSITIVE LOGITS

 best

1.85

best

1.55

 mejor

1.45

 melhor

1.38

 najlepiej

1.34

 bäst

1.32

 BEST

1.27

Best

1.26

最好

1.26

 meilleure

1.24

Activations Density 0.112%

connections related to actions, intentions, and effectiveness within various contexts

best

No Comments

No Known Activations