INDEX

Explanations

programming syntax elements and structure-related terms

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

code closing brace

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

UnusedPrivate

-0.53

msgTypes

-0.51

Vidite

-0.50

Kjelder

-0.49

 nonUne

-0.48

Parcelize

-0.45

IndentedString

-0.45

 ویکی‌پدیا

-0.43

 actor

-0.42

 androgen

-0.40

POSITIVE LOGITS

 CreateTagHelper

0.72

setVerticalGroup

0.58

 CURIAM

0.45

 EconPapers

0.44

#+#

0.44

 NSCoder

0.44

 despe

0.41

 كومونز

0.40

новниш

0.40

 мәкал

0.39

Activations Density 0.030%

programming syntax elements and structure-related terms

code closing brace

No Comments

No Known Activations