INDEX

Explanations

terms related to classification and phases in a structured context

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

the first item

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

/**

-0.48

extAlignment

-0.47

cookieParser

-0.44

MessageTagHelper

-0.44

KommentareTeilen

-0.39

__*/

-0.39

脚注の使い方

-0.39

Bakgrunnsstoff

-0.37

 ब्रेकडाउन

-0.37

новниш

-0.37

POSITIVE LOGITS

енча

0.52

atical

0.51

 ComVisible

0.46

akei

0.46

PLER

0.46

whelming

0.46

Cyto

0.45

oused

0.45

="@+

0.45

ukul

0.45

Activations Density 0.079%

terms related to classification and phases in a structured context

the first item

No Comments

No Known Activations