INDEX

Explanations

patterns related to numerical data or structured sequences

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

hexadecimal values and structural delimiters

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 31-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_31/width_131k/average_l0_109

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.31.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

 betweenstory

-0.59

Ārējās

-0.43

RegressionTest

-0.42

++]=

-0.38

KeyValuePair

-0.37

性

-0.36

()]

-0.36

 Normdatei

-0.35

*]

-0.35

()])

-0.35

POSITIVE LOGITS

 disambiguazione

0.63

ंदीखरीदारी

0.51



0.49

inSlope

0.48

nodoc

0.47

 assoluto

0.47

StoryboardSegue

0.46

 wußte

0.45

printStackTrace

0.45

RTGC

0.45

Activations Density 0.057%

patterns related to numerical data or structured sequences

hexadecimal values and structural delimiters

No Comments

No Known Activations