INDEX

Explanations

phrases that convey understanding or clarity of concepts

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

idea and understanding

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

échange

-0.41

 testé

-0.39

 forskning

-0.37

 obě

-0.37

isamment

-0.37

 spécial

-0.35

 rutas

-0.35

PostMapping

-0.35

 výro

-0.34

non

-0.34

POSITIVE LOGITS

Tikang

0.66

RTGC

0.65

 Überblick

0.65

ंदीखरीदारी

0.63

 disambiguazione

0.63

GraphicsUnit

0.63

 estimate

0.63

 estimating

0.62

SourceChecksum

0.62

 informée

0.61

Activations Density 0.019%

phrases that convey understanding or clarity of concepts

idea and understanding

No Comments

No Known Activations