INDEX

Explanations

specific terms related to techniques, variables, distributions, and actions in technical or scientific contexts

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

group of things

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

 odeur

-0.61

penup

-0.54

 Wikiseite

-0.51

 grammi

-0.48

 leçons

-0.47

 metade

-0.47

 helft

-0.46

 miniaturka

-0.46

 lecciones

-0.46

 leçon

-0.45

POSITIVE LOGITS

parsedMessage

0.48

ImageContext

0.45

ChildScrollView

0.44



0.39

InstanceState

0.38

ranging

0.37

々が

0.37

IVer

0.37

脚注の使い方

0.37

 CURIAM

0.37

Activations Density 0.086%

specific terms related to techniques, variables, distributions, and actions in technical or scientific contexts

group of things

No Comments

No Known Activations