INDEX

Explanations

terms related to non-standard or non-conventional classifications

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

The prefix "non-"

np_acts-logits-general · gemini-2.0-flash

words starting with non

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 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

ScopeManager

-0.60

 Beschluss

-0.50

 Zeitpunkt

-0.49

 betrekking

-0.48

ViewStyle

-0.48

 izquier

-0.48

 Schritt

-0.47

 Erwä

-0.47

 getTitle

-0.47

 îna

-0.46

POSITIVE LOGITS

Non

1.35

non

1.33

Non

1.20

non

1.20

NON

1.10

NON

0.98

 Nons

0.91

 nons

0.85

 Nong

0.82

nons

0.81

Activations Density 0.138%

terms related to non-standard or non-conventional classifications

The prefix "non-"

words starting with non

No Comments

No Known Activations