INDEX

Explanations

references to cultural or historical elements

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

interpretation or meaning

np_acts-logits-general · gemini-2.0-flash

interpretations and clues

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 24-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_24/width_16k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.24.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

brz

-0.32

帳

-0.32

 Halen

-0.32

handleDelete

-0.31

 administr

-0.31

 guarantee

-0.30

 spowod

-0.30

 required

-0.30

 createState

-0.30

Introdução

-0.29

POSITIVE LOGITS

 clues

0.78

interpret

0.75

 hints

0.73

 interpreta

0.70

 interpretations

0.70

 symbolism

0.69

 interpreting

0.69

subtle

0.69

 interpretive

0.69

interpretation

0.68

Activations Density 0.328%

references to cultural or historical elements

interpretation or meaning

interpretations and clues

No Comments

No Known Activations