INDEX

Explanations

references to surveys, nodes, replacements, revisions, images, functions, trials, and messages in a programming context

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

code structures and data types

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

verwijspagina

-0.42

 rahasia

-0.33

 gezien

-0.32

 Gehir

-0.30

 menschliche

-0.29

 Mitarbeit

-0.29

 akibat

-0.28

That

-0.28

 jūs

-0.27

 Unfall

-0.27

POSITIVE LOGITS

IANGLES

0.72

 useAppContext

0.68

AddTagHelper

0.68

UserScript

0.67

HRS

0.66

Бахар

0.66

 surla

0.65

 JpaRepository

0.65

 createSprite

0.65

 CreateTagHelper

0.64

Activations Density 0.129%

references to surveys, nodes, replacements, revisions, images, functions, trials, and messages in a programming context

code structures and data types

No Comments

No Known Activations