INDEX

Explanations

references to the programming language Python

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

Python programming

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

makeConstraints

-0.66

 väh

-0.44

andExpect

-0.43

 dõi

-0.42

 useRef

-0.40

 незавершена

-0.40

refuge

-0.40

fjspx

-0.40

 berücksichtigen

-0.40

isBlank

-0.39

POSITIVE LOGITS

 Python

1.38

Python

1.34

python

1.24

 python

1.17

 PYTHON

1.15

PYTHON

0.93

PyObject

0.78

 PyObject

0.71

썬

0.63

PyExc

0.60

Activations Density 0.006%

references to the programming language Python

Python programming

No Comments

No Known Activations