INDEX

Explanations

occurrences of the word "nome" or related variations

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

nom / name / nombre

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

Cec

-0.62

effect

-0.60

 Galli

-0.59

cruise

-0.56

efficiency

-0.53

Pea

-0.52

 FISCAL

-0.52

שְׁ

-0.52

through

-0.51

effects

-0.51

POSITIVE LOGITS

 name

0.90

Name

0.86

 Name

0.82

 NAME

0.76

NAME

0.76

 Nombre

0.75

 nombre

0.74

name

0.73

 names

0.72

Nombre

0.71

Activations Density 0.002%

occurrences of the word "nome" or related variations

nom / name / nombre

No Comments

No Known Activations