INDEX

Explanations

structural elements and formatting within code snippets

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

programming function returns

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

:✨

-0.64

 betweenstory

-0.49

 Monfieur

-0.47

Ārējās

-0.44

новниш

-0.44

rxjs

-0.42

 typelib

-0.42

onedDateTime

-0.41

 ddelweddau

-0.40

RotationOrder

-0.40

POSITIVE LOGITS

cupertino

0.54

 externi

0.46

WindowConstants

0.45

WithIOException

0.45

atiche

0.44

 earth

0.43

normals

0.43

 autorytatywna

0.43

returns

0.43

 lite

0.42

Activations Density 0.002%

structural elements and formatting within code snippets

programming function returns

No Comments

No Known Activations