INDEX

Explanations

phrases indicating a connection or relationship between subjects

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

"with"

np_acts-logits-general · gemini-2.0-flash

introduces elaboration

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

toMatchSnapshot

-0.51

ones

-0.48

We

-0.48

 anyway

-0.46

-0.45

ts

-0.45

-0.44

TS

-0.43

POSITIVE LOGITS

 lots

1.14

 plenty

1.07

no

1.03

 many

0.91

0.89

 occasional

0.88

 some

0.88

 several

0.86

adpleegd

0.81

 Lots

0.81

Activations Density 0.230%

phrases indicating a connection or relationship between subjects

"with"

introduces elaboration

No Comments

No Known Activations

phrases indicating a connection or relationship between subjects

"with"

introduces elaboration

No Comments

No Known Activations