INDEX

Explanations

references to data structures and their associated operations

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

HDR struct fields

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

 فريبيس

-0.91

WriteTagHelper

-0.88

Autoritní

-0.85

<unused8>

-0.81

<unused52>

-0.81

[@BOS@]

-0.81

<unused28>

-0.81

 चीज़ों

-0.81

<unused14>

-0.80

<unused16>

-0.80

POSITIVE LOGITS

 perna

0.27

 respectively

0.27

笑意

0.26

 respectivamente

0.25

was

0.24

por

0.24

 tendances

0.24

is

0.23

 varandra

0.23

Activations Density 0.045%

references to data structures and their associated operations

HDR struct fields

No Comments

No Known Activations