INDEX

Explanations

terms associated with programming structures and mathematical concepts

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

transistor, calibration, assumption, sanitized, model, votes, conductance, read, digital

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

 ویکی‌پدی

-0.65

GenerationType

-0.58

ValueGeneration

-0.55

RTLD

-0.54

mybatisplus

-0.54

 frattempo

-0.54

mne

-0.53

期刊论文

-0.53

 ProtoMessage

-0.52

Parcelize

-0.52

POSITIVE LOGITS

 surla

0.61

tagHelperRunner

0.54

AutoScaleMode

0.44

httphttps

0.35

 Irán

0.32

BoxShadow

0.31

 anſ

0.30

 poffe

0.29

archiviato

0.29

 resourceCulture

0.29

Activations Density 0.163%

terms associated with programming structures and mathematical concepts

transistor, calibration, assumption, sanitized, model, votes, conductance, read, digital

No Comments

No Known Activations