INDEX

Explanations

code-related constructs and operations involving counters or iterators in programming

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

code structure

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

 surla

-0.57

 rhestr

-0.50

Composable

-0.48

 standing

-0.47

 scre

-0.46

----</

-0.46

 Lombardi

-0.46

 referenties

-0.46

 ProtoMessage

-0.46

CreateModel

-0.45

POSITIVE LOGITS

ImageContext

0.36

ceğine

0.33

eamento

0.33

ryd

0.33

حوالہ

0.33

Попис

0.32

 féminin

0.32

MemoryWarning

0.31

lanations

0.31

 collège

0.31

Activations Density 0.006%

code-related constructs and operations involving counters or iterators in programming

code structure

No Comments

No Known Activations