INDEX

Explanations

phrases indicating lack of effectiveness or limitations in guidelines and strategies

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

preceding a contrast or contradiction

np_acts-logits-general · gemini-2.0-flash

but introduces contrast or exception

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 пусть

-0.58

 niech

-0.52

但不

-0.51

なら

-0.50

也算是

-0.49

λοι

-0.48

BagConstraints

-0.48

 כן

-0.48

也可

-0.46

 gärna

-0.46

POSITIVE LOGITS

 neither

1.01

なかなか

0.99

 none

0.93

neither

0.90

 Unless

0.90

Unless

0.89

 unless

0.89

 lacks

0.88

残念ながら

0.88

 Neither

0.88

Activations Density 0.590%

phrases indicating lack of effectiveness or limitations in guidelines and strategies

preceding a contrast or contradiction

but introduces contrast or exception

No Comments

No Known Activations