INDEX

Explanations

specific numerical values and temporal indicators related to data or predictions

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

quantities and time units

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

modelBuilder

-0.43

Ler

-0.38

 Infórmanos

-0.38

 suport

-0.37

 voie

-0.37

bè

-0.37

URDAY

-0.36

 daire

-0.36

 siku

-0.35

 пер

-0.35

POSITIVE LOGITS

CloseOperation

0.59

 afgelopen

0.46

paravant

0.46

 surla

0.45

+#+#

0.45

 nuova

0.44

 nowej

0.43

 ultimi

0.43

 viime

0.43

 nueva

0.43

Activations Density 0.246%

specific numerical values and temporal indicators related to data or predictions

quantities and time units

No Comments

No Known Activations