INDEX

Explanations

specific patterns related to numerical values, particularly in a programming or technical context

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

docker, ports, deploy

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

en

-0.36

 proven

-0.34

InteropServices

-0.33

XtraEditors

-0.33

Sel

-0.30

addFlags

-0.30

]:

-0.30

ca

-0.30

ρους

-0.30

onames

-0.29

POSITIVE LOGITS

 houſe

0.68

 invokingState

0.68

FunctionFlags

0.66

 autorytatywna

0.65

 snippetHide

0.64

WebServlet

0.63

 Numerade

0.61

 kapturem

0.61

 ModelExpression

0.60

httphttps

0.60

Activations Density 0.002%

specific patterns related to numerical values, particularly in a programming or technical context

docker, ports, deploy

No Comments

No Known Activations