INDEX

Explanations

programming-related elements, particularly in the context of modules and libraries

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

legal citations with WL

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

 BoxFit

-0.66

 wireType

-0.63

DockStyle

-0.62

contentLoaded

-0.60

RenderAtEndOf

-0.59

 MainAxisSize

-0.58

فایل‌لار

-0.57

قایناق‌لار

-0.57

 كومونز

-0.56

 Roskov

-0.56

POSITIVE LOGITS

Хьажоргаш

0.56

graphql

0.53

newInstance

0.49

epar

0.48

PARAM

0.47

js

0.42

ize

0.42

hbs

0.41

izer

0.40

mtr

0.40

Activations Density 0.005%

programming-related elements, particularly in the context of modules and libraries

legal citations with WL

No Comments

No Known Activations