INDEX

Explanations

programming-related terms and concepts associated with navigation and state management in user interfaces

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

actions and changes

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

 raiſ

-0.47

 deſt

-0.46

 quæ

-0.42

 itſelf

-0.37

 ſtill

-0.36

 effectively

-0.36

 sprá

-0.36

 tranſ

-0.35

 gezondheid

-0.35

 âmes

-0.35

POSITIVE LOGITS

(!__

0.62

uxxxx

0.60

tagHelperRunner

0.59

ChildScrollView

0.57

enumii

0.56

principalColumn

0.55

RegressionTest

0.55

 <>",

0.54

bihan

0.51

#+#

0.50

Activations Density 0.057%

programming-related terms and concepts associated with navigation and state management in user interfaces

actions and changes

No Comments

No Known Activations