INDEX

Explanations

positions of authority or leadership within organizations

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

job titles

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

addPreferredGap

-0.61

enetrable

-0.50

ricula

-0.49

띄

-0.49

jména

-0.46

elfde

-0.46

entuk

-0.45

 coupes

-0.45

RegistryLite

-0.45

Derp

-0.45

POSITIVE LOGITS

 houſe

0.41

 UserDao

0.36

 ſon

0.36

 nakalista

0.35

omitempty

0.35

atguigu

0.35

ſelf

0.33

 pleaſure

0.32

UserScript

0.32

 ſch

0.32

Activations Density 0.028%

positions of authority or leadership within organizations

job titles

No Comments

No Known Activations