INDEX

Explanations

references to death or dying situations

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

approaching end

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

SPATH

-0.48

enumii

-0.44

 Roskov

-0.43

EDEFAULT

-0.43

 EconPapers

-0.42

الدراسه

-0.41

 invito

-0.39

Stayed

-0.36

optionalTypeArgs

-0.36

 GoogleFonts

-0.36

POSITIVE LOGITS

 dying

1.38

 nearing

1.27

dying

1.13

 expiring

1.07

 Dying

1.05

 waning

1.03

Dying

0.94

 умира

0.87

 fading

0.86

 approaching

0.85

Activations Density 0.056%

references to death or dying situations

approaching end

No Comments

No Known Activations

references to death or dying situations

approaching end

No Comments

No Known Activations