INDEX

Explanations

concepts related to revenge and retribution

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

seeking revenge

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

 Gefahr

-0.43

 Folge

-0.43

 attravers

-0.42

 ProtoMessage

-0.42

 innamor

-0.40

withIdentifier

-0.39

 dangers

-0.38

 Hälfte

-0.38

 montanha

-0.38

빠

-0.37

POSITIVE LOGITS

 revenge

0.70

revenge

0.61

 venge

0.54

Revenge

0.54

 Revenge

0.53

UserScript

0.53

 retribution

0.51

 vengeance

0.50

 rage

0.50

principalTable

0.50

Activations Density 0.055%

concepts related to revenge and retribution

seeking revenge

No Comments

No Known Activations