INDEX

Explanations

terms related to reduction and minimization in various contexts

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

reducing or minimizing values

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

 gärna

-0.46

SpringRunner

-0.44

GrantedAuthority

-0.43

 propOrder

-0.40

 insuffisamment

-0.40

 riktigt

-0.39

humanité

-0.38

impresa

-0.38

HomeAsUpEnabled

-0.38

 voud

-0.37

POSITIVE LOGITS

 reduc

0.60

 reducing

0.56

 minimizing

0.55

 reduced

0.54

 réduite

0.53

 loss

0.52

 REDUC

0.52

癬

0.51

 minimizar

0.51

 minimized

0.51

Activations Density 0.059%

terms related to reduction and minimization in various contexts

reducing or minimizing values

No Comments

No Known Activations