INDEX

Explanations

phrases that indicate organization and categorization

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Vanity

-0.42

onnaissance

-0.39

iTunes

-0.38

Lalu

-0.37

VAN

-0.37

 argc

-0.37

Lastly

-0.36

 rerum

-0.36

Staff

-0.36

Aux

-0.35

POSITIVE LOGITS

GOTREF

0.57

новништво

0.55

">//

0.49

appartamento

0.48

########.

0.48

operazione

0.47

UnusedPrivate

0.47

 <<<<<<<<<<<<<<

0.45

sofá

0.44

 Había

0.43

Activations Density 0.025%

phrases that indicate organization and categorization

No Comments

No Known Activations