INDEX

Explanations

references to a specific individual named Joseph

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

Joseph Sanders

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

 gwaran

-0.50

back

-0.44

 kontroli

-0.43

at

-0.43

HideInInspector

-0.42

 Tatsache

-0.42

control

-0.41

 übri

-0.41

ta

-0.41

background

-0.41

POSITIVE LOGITS

 Joseph

1.05

Joseph

0.96

 joseph

0.91

 JOSEPH

0.90

JOSEPH

0.86

joseph

0.83

 defaultstate

0.79

 Josephine

0.75

 Josephus

0.72

 Jozef

0.69

Activations Density 0.006%

references to a specific individual named Joseph

Joseph Sanders

No Comments

No Known Activations