INDEX

Explanations

terms related to utility functions or utility in general

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

java utility code

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

tahankan

-0.45

 Gefang

-0.45

extré

-0.42

ckså

-0.42

SpringBootTest

-0.41

 useAppContext

-0.36

UIButton

-0.36

 perched

-0.36

 Gegner

-0.35

 choisissez

-0.35

POSITIVE LOGITS

 initComponents

0.72

 استنادى

0.61

VersionUID

0.61

setViewportView

0.61

 disponibilités

0.60

util

0.60

Datuak

0.55

Rhestr

0.55

RegressionTest

0.54

 snippetHide

0.53

Activations Density 0.000%

terms related to utility functions or utility in general

java utility code

No Comments

No Known Activations