INDEX

Explanations

names of individuals, particularly those associated with academic and research contexts

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

academic titles and 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

########.

-0.52

 ProtoMessage

-0.44

 ModelExpression

-0.34

 دیکھیے

-0.33

 doccia

-0.32

 fotografico

-0.32

 kaarangay

-0.31

Rüyada

-0.31

følgelig

-0.31

MigrationBuilder

-0.30

POSITIVE LOGITS

 Professor

0.69

professor

0.65

 professor

0.61

 profesora

0.60

 PROFESSOR

0.58

Professor

0.58

教授

0.58

 Professors

0.58

SuppressLint

0.54

Expertise

0.54

Activations Density 0.028%

names of individuals, particularly those associated with academic and research contexts

academic titles and names

No Comments

No Known Activations