INDEX

Explanations

code or programming language elements, especially related to libraries and function definitions

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

ric followed by ator

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

-0.44

-0.35

ci

-0.35

st

-0.34

 orejas

-0.33

ity

-0.33

-0.32

POSITIVE LOGITS

 Roskov

1.01

 informée

1.00

 Italijanski

0.95

 Numerade

0.91

 queſta

0.90

 Италијани

0.89

 للاسماء

0.88

 Meksiku

0.87

хьтан

0.85

iſen

0.84

Activations Density 0.021%

code or programming language elements, especially related to libraries and function definitions

ric followed by ator

No Comments

No Known Activations