INDEX

Explanations

values and their relationships in mathematical expressions

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

options following parentheses

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

感

-0.35

<eos>

-0.33

不可

-0.31

uxley

-0.30

Còn

-0.29

λ

-0.29

buie

-0.28

rstrip

-0.28

 superb

-0.28

δά

-0.28

POSITIVE LOGITS

 betweenstory

0.62

ロウィン

0.61

 يتيمه

0.59

StoryboardSegue

0.58

 שוליים

0.58

 autorytatywna

0.57

majánló

0.56

 Infórmanos

0.53

ſicht

0.53

XmlAccessType

0.52

Activations Density 0.244%

values and their relationships in mathematical expressions

options following parentheses

No Comments

No Known Activations