INDEX

Explanations

patterns related to JSON object structures and their properties

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

field/attribute names

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

 cucharadas

-0.33

 campamento

-0.30

 seguramente

-0.29

 mantenido

-0.28

又

-0.28

 pasada

-0.27

 unless

-0.27

罢

-0.26

 thus

-0.26

 peuples

-0.26

POSITIVE LOGITS

 CreateTagHelper

0.73

 nakalista

0.72

ftagPool

0.69

 nonUne

0.69

 للمعارف

0.69

adaptiveStyles

0.66

queryInterface

0.66

 propOrder

0.64

[@"

0.64

 Administrativna

0.63

Activations Density 0.015%

patterns related to JSON object structures and their properties

field/attribute names

No Comments

No Known Activations