INDEX

Explanations

code-related terms and structures in programming, particularly those associated with classifiers and predictive functions

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

feature selection methods

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

Merkmale

-0.53

 désolés

-0.49

 nakalista

-0.49

+#+

-0.49

HtmlAttribute

-0.48

 faſt

-0.46

interopRequire

-0.45

 empre

-0.44

adaptiveStyles

-0.43

 propOrder

-0.43

POSITIVE LOGITS

GTCX

0.42

WriteBarrier

0.40

 Anhäng

0.38

RefNanny

0.38

 цело

0.37

 ля

0.34

 ExecuteAsync

0.33

 Weisheit

0.33

Stayed

0.32

 הל

0.32

Activations Density 0.224%

code-related terms and structures in programming, particularly those associated with classifiers and predictive functions

feature selection methods

No Comments

No Known Activations