INDEX

Explanations

phrases indicating a reply or reaction related to requests or inquiries

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

in response to

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

 Constitucional

-0.59

 constitucional

-0.50

eterangan

-0.47

 administratif

-0.45

 Constitutional

-0.45

MessageOf

-0.44

AllAfrica

-0.43

 institucional

-0.43

 mensaj

-0.42

 Rockefeller

-0.42

POSITIVE LOGITS

__(/*!

0.64

 estekak

0.56

dientemente

0.50

 behalf

0.49

 sumpay

0.47

CascadeType

0.47

OGND

0.45

 accordance

0.44

 spite

0.44

Rüyada

0.44

Activations Density 0.045%

phrases indicating a reply or reaction related to requests or inquiries

in response to

No Comments

No Known Activations