INDEX

Explanations

code structure elements like classes, methods, and their access modifiers

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

user input

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

httphttps

-0.44

ėjimas

-0.43

nezeu

-0.42

+#+#

-0.41

tener

-0.40

ientras

-0.40

اصله

-0.40

достатки

-0.40

pots

-0.40

Carriera

-0.39

POSITIVE LOGITS

 kaarangay

0.54

цездатний

0.49

GEBURTSDATUM

0.44

fromnode

0.41

 OFDb

0.41

feitura

0.41

Географиясе

0.40

ウンサー

0.40

Derbyniad

0.39

istoitu

0.39

Activations Density 0.002%

code structure elements like classes, methods, and their access modifiers

user input

No Comments

No Known Activations