INDEX

Explanations

concepts related to mathematical systems and structures

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

increments, parts, and root systems

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

 Schilder

-0.36

hed

-0.35

 consultato

-0.35

 bản

-0.33

生

-0.32

 Просто

-0.32

geordnet

-0.31

 Schar

-0.31

복

-0.31

han

-0.31

POSITIVE LOGITS

 Bewußt

0.57

 vrijwilli

0.53

genodigd

0.52

acijos

0.50

 løs

0.50

 vermogen

0.49

 виправивши

0.48

पया

0.47

IsMutable

0.47

 perdere

0.47

Activations Density 0.207%

concepts related to mathematical systems and structures

increments, parts, and root systems

No Comments

No Known Activations