INDEX

Explanations

mathematical or computational terms related to matrices

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

matrix then closing brace

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

RegressionTest

-0.48

Kamer

-0.43

isamment

-0.43

orianCalendar

-0.41

orded

-0.41

othe

-0.41

kjø

-0.41

ONIA

-0.41

errHandler

-0.41

nthesis

-0.41

POSITIVE LOGITS

PrototypeOf

0.61

Statics

0.60

__:

0.59

IUrlHelper

0.54

stateProvider

0.49

StoryboardSegue

0.45

writerow

0.43

 Roskov

0.43

Tikang

0.42

Activations Density 0.000%

mathematical or computational terms related to matrices

matrix then closing brace

No Comments

No Known Activations