INDEX

Explanations

terms related to breast cancer and its detection

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

Junk, filler words, or placeholders

np_acts-logits-general · gemini-2.0-flash

end array, sample

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 28-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_28/width_16k/average_l0_119

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.28.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

 Additionally

-0.61

 additionally

-0.61

 sahiptir

-0.58

 tevens

-0.57

 insbesondere

-0.57

であるが

-0.55

 następnie

-0.54

 subsequently

-0.54

Additionally

-0.54

 Subsequently

-0.53

POSITIVE LOGITS

 thingy

1.13

 stuff

1.02

 crappy

0.98

 weird

0.94

 freakin

0.94

 dudes

0.93

 goodies

0.93

 kinda

0.88

 blah

0.86

 nasty

0.85

Activations Density 1.262%

terms related to breast cancer and its detection

Junk, filler words, or placeholders

end array, sample

No Comments

No Known Activations