INDEX

Explanations

references to programming classes and components in code

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

module, runner, agent, snow, help

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

enumi

-0.56

ьаж

-0.49

__*/

-0.47

enumii

-0.46

assemblée

-0.45

 transfieras

-0.42

oredCriteria

-0.42

featureID

-0.42

OGND

-0.42

*~*~

-0.42

POSITIVE LOGITS

expandindo

0.58

AutoScaleMode

0.44

SpringBootTest

0.44

bewerken

0.42

 Réponses

0.42

PointerException

0.38

 modulation

0.38

MessageOf

0.37

nestjs

0.37

 дописавши

0.37

Activations Density 0.016%

references to programming classes and components in code

module, runner, agent, snow, help

No Comments

No Known Activations