INDEX

Explanations

key-value pairs and structured data formats

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

CloseOperation

-0.84

 betweenstory

-0.77

 nahilalakip

-0.76

 informée

-0.76

 propOrder

-0.73

 autorytatywna

-0.72

Personendaten

-0.71

enumii

-0.71

FunctionFlags

-0.71

MLLoader

-0.69

POSITIVE LOGITS

同

0.39

).

0.38

0.36

再

0.34

..

0.34

 mismas

0.33

 ésta

0.33

 samen

0.32

 autonomía

0.32

'.

0.32

Activations Density 0.011%

key-value pairs and structured data formats

No Comments

No Known Activations