INDEX

Explanations

references to programming actions and functions in coding contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 2-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_2/width_16k/average_l0_67

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.2.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

itinéraire

-0.39

act

-0.36

 Menschheit

-0.35

SBATCH

-0.35

 typelib

-0.34

RefreshLayout

-0.34

 schlafen

-0.33

 Políticas

-0.32

RegressionTest

-0.32

ant

-0.31

POSITIVE LOGITS

Action

1.96

 Action

1.85

 Actions

1.13

 ACTION

0.97

 Acción

0.96

Actions

0.94

Active

0.93

 Activ

0.93

 Active

0.91

 Activity

0.89

Activations Density 0.335%

references to programming actions and functions in coding contexts

No Comments

No Known Activations