INDEX

Explanations

evidence of data presence and significance in various studies

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

finding confirmed or found

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

vician

-0.57

thâu

-0.48

mitian

-0.45

 queſta

-0.45

cæ

-0.44

出版年

-0.43

 ſch

-0.41

AddTagHelper

-0.41

 houſe

-0.40

 pous

-0.40

POSITIVE LOGITS

 dabei

0.54

SourceChecksum

0.52

RenderAtEndOf

0.51

原始内容存档

0.49

原始内容存档于

0.45

脚注の使い方

0.44

UrlResolution

0.43

rungsseite

0.41

 Dabei

0.40

TLDR

0.40

Activations Density 0.166%

evidence of data presence and significance in various studies

finding confirmed or found

No Comments

No Known Activations