INDEX

Explanations

references to APIs and their related exceptions or classes

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

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

vaadin

-0.41

onViewCreated

-0.38

PyTuple

-0.38

 akar

-0.36

ViewFeatures

-0.35

 Baghdad

-0.34

MUD

-0.34

saraba

-0.34

sap

-0.33

AddTagHelper

-0.33

POSITIVE LOGITS

 simplifié

0.61

 HFILL

0.52

 Италијани

0.51

 utafitiHapana

0.51

simplified

0.48

 sää

0.48

Simplified

0.47

 Layanan

0.47

拗

0.45

 Präsidenten

0.45

Activations Density 0.069%

references to APIs and their related exceptions or classes

No Comments

No Known Activations