INDEX

Explanations

sentences that suggest recommendations or advice

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

Kariera

-0.45

 opdracht

-0.42

 syke

-0.40

Gemeinden

-0.39

Muri

-0.38

Ứ

-0.38

 Kraw

-0.38

PathVariable

-0.38

 Stodd

-0.37

Zuf

-0.37

POSITIVE LOGITS

 advisable

0.57

 Caution

0.54

 warto

0.54

always

0.52

 always

0.50

 слід

0.50

advised

0.50

 sebaiknya

0.49

 conviene

0.49

 caution

0.48

Activations Density 0.024%

sentences that suggest recommendations or advice

No Comments

No Known Activations