INDEX

Explanations

references to problems, investigations, and the process of addressing issues

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

break down or declared incident

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

 enseñanza

-0.31

 popularidad

-0.30

เค้า

-0.27

phylla

-0.27

 passione

-0.27

 passion

-0.27

很多的

-0.26

uli

-0.25

 desej

-0.24

CreateModel

-0.24

POSITIVE LOGITS

RegressionTest

0.70

ThroughAttribute

0.70

 OFDb

0.69

}{@

0.61

SequentialGroup

0.60



0.59

 disambiguazione

0.59

!*\

0.58

 NSCoder

0.57

emergency

0.57

Activations Density 0.017%

references to problems, investigations, and the process of addressing issues

break down or declared incident

No Comments

No Known Activations