INDEX

Explanations

mathematical symbols and operations related to equations and expressions

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

numbers in equations

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

 nakalista

-0.86

RegressionTest

-0.86

 AssemblyTitle

-0.82

rungsseite

-0.82

 kasarigan

-0.81

iſchen

-0.76

principalColumn

-0.75

<unused79>

-0.73

<unused41>

-0.73

<unused8>

-0.73

POSITIVE LOGITS

치

0.25

kol

0.23

↵↵

0.21

kov

0.21

EIF

0.21

玉

0.21

 सं

0.21

('/')

0.21

Chi

0.20

tip

0.20

Activations Density 0.365%

mathematical symbols and operations related to equations and expressions

numbers in equations

No Comments

No Known Activations