INDEX

Explanations

syntax-related elements in programming code

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

early return in code blocks

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

 modelled

-0.33

kol

-0.29

bootstrapcdn

-0.29

兼

-0.29

 reported

-0.29

CLR

-0.29

着

-0.28

 flagged

-0.28

kol

-0.28

 coated

-0.28

POSITIVE LOGITS

 للاسماء

0.93

 فريبيس

0.75

principalColumn

0.72

#+#

0.70

enderror

0.70

دانشنامهٔ

0.68

 CreateTagHelper

0.66

WaitGroup

0.66

Personensuche

0.65

AnchorStyles

0.63

Activations Density 0.009%

syntax-related elements in programming code

early return in code blocks

No Comments

No Known Activations