INDEX

Explanations

sections of code or programming language constructs

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

programming and scientific content

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

BagConstraints

-0.47

 Komunikasi

-0.41

 anyways

-0.41

 inoxidable

-0.40

 Anyways

-0.40

ähteet

-0.39

 onor

-0.39

antaranya

-0.39

jména

-0.38

Anyways

-0.38

POSITIVE LOGITS

 EdgeInsets

0.66

BuildContext

0.55

WireFormatLite

0.55

Autoritní

0.54

Географиясе

0.51

nodoc

0.50

 Iowa

0.48

flutter

0.46

 Covid

0.45

hasPermission

0.45

Activations Density 0.001%

sections of code or programming language constructs

programming and scientific content

No Comments

No Known Activations