INDEX

Explanations

phrases that indicate comparisons and evaluations, particularly regarding quality and capabilities

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

"is" followed by a preposition or adjective

np_acts-logits-general · gemini-2.0-flash

is expressed or is devoted

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ConstraintMaker

-0.51

 mida

-0.49

sola

-0.47

mita

-0.46

 Terminal

-0.46

awtextra

-0.46

richten

-0.45

XIM

-0.44

 website

-0.44

 Circle

-0.43

POSITIVE LOGITS

 defaultstate

0.78

 onPostExecute

0.63

="@+

0.63

 onOptions

0.62

nologue

0.60

iotensin

0.59

StructEnd

0.59

Distribuzione

0.59

expandindo

0.58

 wireType

0.58

Activations Density 0.370%

phrases that indicate comparisons and evaluations, particularly regarding quality and capabilities

"is" followed by a preposition or adjective

is expressed or is devoted

No Comments

No Known Activations