INDEX

Explanations

sections of text that discuss programming or code-related structures

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

Code snippets or mathematical notation

np_acts-logits-general · gemini-2.0-flash

closing parenthesis with trailing characters

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-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

 orejas

-0.54

-0.51

-0.46

↵

-0.45

-0.42

、

-0.41

 Tiefen

-0.41

ly

-0.40

al

-0.40

POSITIVE LOGITS

rungsseite

1.19

 queſta

1.17

GEBURTSDATUM

1.14

 незавершена

1.13

 estekak

1.12

 Roskov

1.12

parsedMessage

1.11

OGND

1.08

Хьажоргаш

1.08

تقاوى

1.07

Activations Density 1.079%

sections of text that discuss programming or code-related structures

Code snippets or mathematical notation

closing parenthesis with trailing characters

No Comments

No Known Activations