INDEX

Explanations

terms related to persuasion and influence

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

persuade, persuasion, persuasive

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

crí

-0.45

 manquer

-0.45

 horrible

-0.44

 incompatibility

-0.43

 Lingkungan

-0.43

Hig

-0.43

CFC

-0.42

 horrors

-0.42

 Picchu

-0.41

-0.40

POSITIVE LOGITS

 persuade

1.72

 persuaded

1.54

 persuading

1.41

 persuasion

1.34

 persu

1.34

 persua

1.32

 persuasive

1.21

uasion

1.21

 Pers

0.91

Pers

0.82

Activations Density 0.006%

terms related to persuasion and influence

persuade, persuasion, persuasive

No Comments

No Known Activations