INDEX

Explanations

key moments and actions related to events or announcements involving individuals

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

verbs followed by events or accomplishments

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

 nødven

-0.31

 sitä

-0.28

 nødvendig

-0.28

 juos

-0.28

jį

-0.27

 agujas

-0.27

 vsak

-0.27

 svě

-0.26

 richtet

-0.26

 zicht

-0.25

POSITIVE LOGITS

 ModelExpression

0.82

 surla

0.74

+#+

0.73

MLLoader

0.73

IUrlHelper

0.71

tvguidetime

0.70

EndContext

0.70

parsedMessage

0.70

himself

0.68

 himself

0.67

Activations Density 0.116%

key moments and actions related to events or announcements involving individuals

verbs followed by events or accomplishments

No Comments

No Known Activations