INDEX

Explanations

terms indicating additional information or notes

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

"Additional" followed by a noun

np_acts-logits-general · gemini-2.0-flash

additional facts, files, or items

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

 selanjutnya

-0.60

 also

-0.60

also

-0.59

 demais

-0.59

 novamente

-0.58

 restantes

-0.58

 énergé

-0.57

 berikutnya

-0.56

 rispet

-0.56

 essentiellement

-0.54

POSITIVE LOGITS

 betweenstory

0.98

setVerticalGroup

0.85

 tartalomajánló

0.76

ctors

0.74

 rounds

0.74

worldly

0.74

agus

0.73

ctions

0.73

rungsseite

0.72

 MainAxisSize

0.72

Activations Density 0.206%

terms indicating additional information or notes

"Additional" followed by a noun

additional facts, files, or items

No Comments

No Known Activations