INDEX

Explanations

elements related to documentation and referencing format

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

parts of legal descriptions

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

WriteTagHelper

-0.41

ImageContext

-0.39

cupertino

-0.39

 Usaha

-0.39

Élet

-0.35

 Herbst

-0.35

 Jenkins

-0.34

RegressionTest

-0.34

 Sommers

-0.33

">//

-0.32

POSITIVE LOGITS

ScopeManager

0.70

 Signalez

0.69

 ویکی‌پدی

0.68

تقاوى

0.59

oredCriteria

0.53

PerformLayout

0.53

rungsseite

0.52

 EconPapers

0.51

сылкі

0.50

 queſta

0.50

Activations Density 0.001%

elements related to documentation and referencing format

parts of legal descriptions

No Comments

No Known Activations