INDEX

Explanations

significant numerical or data presentation elements within a text

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

uses, findings, specific items

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

+#+

-0.47

parsedMessage

-0.45

 חיצוניים

-0.44

keyColumn

-0.44

principalTable

-0.43

 herido

-0.38

 oreja

-0.37

 وتسجيلات

-0.36

 calculées

-0.36

*/),

-0.36

POSITIVE LOGITS

 nakalista

0.58

:✨

0.52

 Photocase

0.52

 stanovnika

0.50

 للاسماء

0.48

 kasarigan

0.47

ंदीखरीदारी

0.46

商品説明

0.44

Portale

0.44

ransition

0.42

Activations Density 0.001%

significant numerical or data presentation elements within a text

uses, findings, specific items

No Comments

No Known Activations