INDEX

Explanations

terms related to feature extraction and representation learning in neural networks

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

word embeddings and information

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

setupUi

-0.60

!*\

-0.60

zzleHttp

-0.56

findpost

-0.55

createServer

-0.51

 metav

-0.51

onedDateTime

-0.49

awtextra

-0.49

 elseif

-0.49

pyplot

-0.48

POSITIVE LOGITS

embeddings

0.47

 esternos

0.42

 dignidad

0.39

Distribución

0.36

 beina

0.35

 rehabilitación

0.35

 виправивши

0.34

uldade

0.33

цездатний

0.33

embedding

0.32

Activations Density 0.007%

terms related to feature extraction and representation learning in neural networks

word embeddings and information

No Comments

No Known Activations