INDEX

Explanations

references to collections or aggregations of objects

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

pile and accumulations

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

RegressionTest

-0.78

AndroidJUnit

-0.63

 AppCompatTheme

-0.54

期刊论文

-0.52

 Weib

-0.50

 UnityEngine

-0.49

RuleContext

-0.48

OGND

-0.48

toprule

-0.48

 Astra

-0.48

POSITIVE LOGITS

 pile

2.19

 Pile

2.03

pile

2.02

Pile

2.00

 piles

1.88

piles

1.59

 Piles

1.52

 piled

1.50

 piling

1.49

PILE

1.27

Activations Density 0.003%

references to collections or aggregations of objects

pile and accumulations

No Comments

No Known Activations