INDEX

Explanations

adjectives and terms related to effectiveness and health

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

| irregular | semantically | negative | |

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

 Мексичка

-0.72

DockStyle

-0.67

 تضيفلها

-0.60

хьтан

-0.59

 للمعارف

-0.59

 pinulongan

-0.58

MigrationBuilder

-0.58

 Biôgrafia

-0.56

 kaarangay

-0.56

 vícti

-0.56

POSITIVE LOGITS

 irregular

0.57

 unfit

0.54

 unhealthy

0.53

 inefficient

0.52

 invalid

0.51

 inactive

0.49

 unsuitable

0.49

 Irregular

0.48

 inappropriate

0.48

 incompetent

0.48

Activations Density 0.235%

adjectives and terms related to effectiveness and health

| irregular | semantically | negative | |

No Comments

No Known Activations