INDEX

Explanations

references to sections and results in a document, particularly in relation to numerical data and methodologies

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

section markers

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

+#+

-0.55

 ſch

-0.53

Архівовано

-0.51

 iprot

-0.50

aarrggbb

-0.49

 ſehr

-0.48

PIC

-0.48

iſen

-0.48

StructEnd

-0.48

Personensuche

-0.47

POSITIVE LOGITS

 nahilalakip

0.48

Демографія

0.34

Abitanti

0.34

 <<<<<<<<<<<<<<

0.32

 jembatan

0.32

principalTable

0.32

Espèce

0.31

 Económica

0.31

 discussion

0.31

hyrchwyd

0.30

Activations Density 0.000%

references to sections and results in a document, particularly in relation to numerical data and methodologies

section markers

No Comments

No Known Activations