INDEX

Explanations

symbols and numerical values related to tensor operations or dimensions

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

tensor shapes and dimensions

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

 CreateTagHelper

-0.55

AddTagHelper

-0.51

 surla

-0.51

IContainer

-0.51

AsUp

-0.51

 IndexPath

-0.48

 createState

-0.47

RuleContext

-0.45

findpost

-0.45

 referenties

-0.44

POSITIVE LOGITS

MessageOf

0.46

SHAPE

0.40

 oorsp

0.39

 wijk

0.39

 öffnungszeiten

0.39

 ontwerp

0.38

êques

0.38

 Lokasi

0.38

lacable

0.37

 Marlow

0.37

Activations Density 0.018%

symbols and numerical values related to tensor operations or dimensions

tensor shapes and dimensions

No Comments

No Known Activations