INDEX

Explanations

references to data structures and their properties in code

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

object property assignments

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

 ujednoznacz

-0.52

flashdata

-0.42

onOptions

-0.41

Hamb

-0.41

accordo

-0.41

argout

-0.40

amerikanischer

-0.40

tvguidetime

-0.40

Harm

-0.40

 Hamb

-0.40

POSITIVE LOGITS

elemField

0.50

igshid

0.46

richTextPanel

0.45

awaiter

0.43

 ProtoMessage

0.42

oneofs

0.41

ButterKnife

0.40

Hentet

0.38

 findViewById

0.38

RTHOOK

0.36

Activations Density 0.013%

references to data structures and their properties in code

object property assignments

No Comments

No Known Activations