INDEX

Explanations

words related to rewards and incentivization

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 0-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_0/width_16k/average_l0_105

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.0.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

egrave

-0.70

ParallelGroup

-0.63

EndInit

-0.61

belline

-0.60

UNS

-0.59

rzost

-0.59

endphp

-0.59

culate

-0.59

UnusedPrivate

-0.59

IDENCE

-0.58

POSITIVE LOGITS

 rewards

0.68

Pad

0.60

 gând

0.58

rewards

0.58

 seeds

0.58

Emb

0.57

Emb

0.57

 kohdetta

0.57

search

0.56

 Rewards

0.54

Activations Density 0.068%

words related to rewards and incentivization

No Comments

No Known Activations