INDEX

Explanations

references to duality or dual features in various contexts

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

dual modifier

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-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

 forKey

-0.52

สึก

-0.49

semantics

-0.49

mby

-0.48

berg

-0.47

peche

-0.47

CONSIDER

-0.47

efeller

-0.47

 Lorena

-0.46

PreInfinity

-0.45

POSITIVE LOGITS

Dual

1.03

 dual

1.02

 Dual

1.01

 DUAL

1.00

dual

0.89

 twin

0.75

 Duel

0.71

 duel

0.70

双

0.69

Dua

0.68

Activations Density 0.009%

references to duality or dual features in various contexts

dual modifier

No Comments

No Known Activations