INDEX

Explanations

terms related to software and programming interfaces

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

parts, components, or specific items

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

 Италијани

-0.83

IUrlHelper

-0.78

IndentedString

-0.71

ValueStyle

-0.68

 pinulongan

-0.68

HasAnnotation

-0.65

ніципалі

-0.60

rungsseite

-0.59

localctx

-0.58

adaptiveStyles

-0.56

POSITIVE LOGITS

TextWatcher

0.28

 actionMode

0.25

нига

0.24

interrup

0.24

 elettrica

0.23

CodedInputStream

0.23

 kekerasan

0.23

ledes

0.23

 marmor

0.23

缺点

0.23

Activations Density 0.035%

terms related to software and programming interfaces

parts, components, or specific items

No Comments

No Known Activations