INDEX

Explanations

technical terms and code-related components

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

UI code and bindings

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

待

-0.34

 régle

-0.33

UnusedPrivate

-0.32

 ناش

-0.31

JdbcTemplate

-0.31

ýl

-0.29

 gustó

-0.28

例文

-0.28

Diwedd

-0.27

 preguntó

-0.26

POSITIVE LOGITS

WebServlet

0.70

parsedMessage

0.60

 }}"></

0.59

évaluateur

0.57

ResumeLayout

0.56

XmlAccessorType

0.55

niſſe

0.53

MemoryWarning

0.53

ikusbot

0.52

InputBorder

0.52

Activations Density 0.166%

technical terms and code-related components

UI code and bindings

No Comments

No Known Activations