INDEX

Explanations

terms related to programming and technical specifications

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

technical identifiers

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

 arşivlendi

-0.73

NUMX

-0.73

CPtr

-0.65

ValueStyle

-0.63

nữ

-0.60

EDEFAULT

-0.59

StoryboardSegue

-0.57

 متعلقه

-0.56

сылкі

-0.56

ClientSize

-0.55

POSITIVE LOGITS

/**

0.39

 amarillo

0.37

 izan

0.36

 bağ

0.35

nagem

0.35

 hydrauli

0.34

IRQn

0.34

 geist

0.33

 Governo

0.33

rodow

0.33

Activations Density 0.144%

terms related to programming and technical specifications

technical identifiers

No Comments

No Known Activations