INDEX

Explanations

phrases that introduce examples or clarifications

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 7-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_7/width_16k/average_l0_92

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.7.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

parsedMessage

-0.71

principalColumn

-0.70

новниш

-0.67

 Савезне

-0.65

ftagPool

-0.63

IVEREF

-0.63

tvguidetime

-0.63

 pinulongan

-0.63

OGND

-0.62

InjectAttribute

-0.60

POSITIVE LOGITS

 Which

0.49

Which

0.47

Who

0.37

CrossRef

0.35

Who

0.34

Where

0.34

Or

0.34

ConstraintMaker

0.33

Gdzie

0.33

ymce

0.33

Activations Density 0.693%

phrases that introduce examples or clarifications

No Comments

No Known Activations