INDEX

Explanations

specific nouns and terms related to programming and data analysis

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

organization, festival, differences, viruses, exchange, dir, url

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

DebuggerNonUser

-0.48

 esternos

-0.45

 zufolge

-0.41

 piemē

-0.40

TargetException

-0.40

 désolés

-0.40

andExpect

-0.39

Климаты

-0.38

AnimationsModule

-0.38

BorderFactory

-0.37

POSITIVE LOGITS

openzeppelin

0.54

 betweenstory

0.47

 CreateTagHelper

0.44

NEST

0.43

RL

0.41

 bedside

0.41

$_['

0.40

 Mite

0.40

 diec

0.40

曜

0.40

Activations Density 0.374%

specific nouns and terms related to programming and data analysis

organization, festival, differences, viruses, exchange, dir, url

No Comments

No Known Activations