INDEX

Explanations

references to "hitting" or "strikes" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 15-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_15/width_16k/average_l0_78

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.15.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

┈┈

-0.49

 esboço

-0.49

дым

-0.49

categoryName

-0.48

 rzecz

-0.47

sprozess

-0.47

 vigilancia

-0.47

 boletos

-0.46

Procedimiento

-0.45

🔹

-0.45

POSITIVE LOGITS

hit

1.64

Hit

1.60

 hits

1.58

 hitting

1.58

Hit

1.51

HIT

1.46

hit

1.45

 Hits

1.43

hitting

1.41

HIT

1.38

Activations Density 0.103%

references to "hitting" or "strikes" in various contexts

No Comments

No Known Activations