INDEX

Explanations

terms related to software development and programming libraries

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

contractions and code literals

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

 NSCoder

-0.88

 ब्रेकडाउन

-0.73

 invokingState

-0.71

 queſta

-0.71

 AssemblyCulture

-0.71

KEYCODE

-0.71

 desmotivaciones

-0.71

DockStyle

-0.71

 متعلقه

-0.71

InjectAttribute

-0.68

POSITIVE LOGITS

 operator

0.30

0.28

 hors

0.28

 acquired

0.28

0.27

fa

0.27

 answer

0.26

 союз

0.26

mod

0.26

 куп

0.26

Activations Density 0.042%

terms related to software development and programming libraries

contractions and code literals

No Comments

No Known Activations