INDEX

Explanations

occurrences of complex scientific or mathematical terminology related to results and observations in research

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

scientific or technical terminology describing results, processes, or associated attributes.

oai_token-act-pair · claude-3-5-haiku-20241022 Triggered by @neilrathi

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_20/width_16k/average_l0_68

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

 auroit

-0.77

 pouvoit

-0.69

 feroit

-0.65

了嗎

-0.61

 avoient

-0.59

 ainfi

-0.58

 étoit

-0.57

 étoient

-0.57

Portály

-0.56

 ſon

-0.52

POSITIVE LOGITS

 ongoing

0.50

 meagre

0.46

 tiny

0.44

 pesky

0.44

__*/

0.44

 popular

0.44

 ensuing

0.44

된

0.44

 joint

0.43

"]));

0.43

Activations Density 0.454%

occurrences of complex scientific or mathematical terminology related to results and observations in research

scientific or technical terminology describing results, processes, or associated attributes.

No Comments

No Known Activations