INDEX

Explanations

phrases that indicate measurement or evaluation of significance and authority

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

weight and extent

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

 AttributeSet

-0.58

CodeAttribute

-0.53

 fieldNum

-0.50

SharedDtor

-0.49

errHandler

-0.49

UnknownFieldSet

-0.48

disposing

-0.48

parsedMessage

-0.47

OpenHelper

-0.47

addCriterion

-0.47

POSITIVE LOGITS

 brunt

0.50

Suara

0.43

thal

0.40

 Severity

0.38

läg

0.38

rami

0.37

CollectionUtils

0.37

Force

0.36

 Hancock

0.36

 scraping

0.36

Activations Density 0.040%

phrases that indicate measurement or evaluation of significance and authority

weight and extent

No Comments

No Known Activations