INDEX

Explanations

key concepts related to evidence and its credibility

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

preceding "is" or "are"

np_acts-logits-general · gemini-2.0-flash

concept nouns followed by 'is'

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

도록

-0.55

算了

-0.52

 rues

-0.52

Has

-0.51

yet

-0.49

 match

-0.49

まるで

-0.48

 coincidence

-0.48

かったので

-0.47

commons

-0.47

POSITIVE LOGITS

 typically

0.79

と聞

0.79

 needn

0.78

can

0.75

seamnă

0.75

 generally

0.74

 usually

0.72

≠

0.72

 kän

0.71

ValueStyle

0.71

Activations Density 0.772%

key concepts related to evidence and its credibility

preceding "is" or "are"

concept nouns followed by 'is'

No Comments

No Known Activations