INDEX

Explanations

technical references and dependencies related to programming or software development

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

programming code constructs

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

 nakalista

-0.48

 outside

-0.31

 zám

-0.28

 būt

-0.28

 only

-0.28

 near

-0.27

 zugelassen

-0.27

 around

-0.27

 børn

-0.26

 Kell

-0.26

POSITIVE LOGITS

IUrlHelper

0.94

 AssemblyCulture

0.92

 BrowserModule

0.77

WithIOException

0.73

InjectAttribute

0.69

 CreateTagHelper

0.67

Rujuakan

0.66

ientras

0.64

 Verſ

0.60

LookAnd

0.59

Activations Density 0.017%

technical references and dependencies related to programming or software development

programming code constructs

No Comments

No Known Activations