INDEX

Explanations

numerical references and statistics within the text

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

find the

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

httphttps

-0.57

 &___

-0.51

ArrowToggle

-0.51

 Савезне

-0.51

 Numerade

-0.49

tagHelperRunner

-0.49

馃

-0.45

цездатний

-0.44

 transfieras

-0.44

 pinulongan

-0.43

POSITIVE LOGITS

 tardes

0.45

 pozdě

0.45

 splet

0.43

 acompaña

0.42

 držav

0.41

 piacere

0.41

 Rücken

0.41

 picioare

0.40

 acompañada

0.40

 acompañado

0.40

Activations Density 0.287%

numerical references and statistics within the text

find the

No Comments

No Known Activations