INDEX

Explanations

programming-related syntax and constructs for table view data handling

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

code structure and syntax

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

featureID

-0.71

parsedMessage

-0.66

AndEndTag

-0.66

WebServlet

-0.62

addCriterion

-0.59

elemField

-0.59

FormTagHelper

-0.59

XMLSchema

-0.57

CloseOperation

-0.56

GenerationType

-0.55

POSITIVE LOGITS

 abiye

0.37

bildēt

0.36

 vuel

0.35

 NEED

0.34

 fiecare

0.34

Forever

0.34

indah

0.33

 Tiefen

0.33

nisk

0.33

idste

0.33

Activations Density 0.001%

programming-related syntax and constructs for table view data handling

code structure and syntax

No Comments

No Known Activations