INDEX

Explanations

terms related to finishing actions or outcomes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 8-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_8/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.8.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

bytes

-0.78

pyr

-0.76

MigrationBuilder

-0.76

ROLE

-0.73

 Krä

-0.71

Coyote

-0.70

triangleq

-0.70

лтамалар

-0.70

 AppModule

-0.68

tritte

-0.68

POSITIVE LOGITS

 FINISH

1.45

 Finish

1.44

 finish

1.42

 finished

1.36

finish

1.32

 finishes

1.32

Finish

1.32

FINISH

1.29

 Finished

1.28

 finishing

1.23

Activations Density 0.075%

terms related to finishing actions or outcomes

No Comments

No Known Activations