INDEX

Explanations

references to patches or updates in a technical context

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

patch

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

stdlib

-0.53

Autoritní

-0.51

iostream

-0.42

OGND

-0.42

umato

-0.41

 erwartet

-0.40

esModule

-0.40

 createUser

-0.40

 NavController

-0.39

hasErrors

-0.39

POSITIVE LOGITS

 Patch

1.49

 patch

1.48

patch

1.41

Patch

1.38

 PATCH

1.35

 patches

1.20

 Patches

1.20

 patching

1.20

 patched

1.09

PATCH

1.09

Activations Density 0.005%

references to patches or updates in a technical context

patch

No Comments

No Known Activations