INDEX

Explanations

expressions indicating personal experience or attempts at problem-solving

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

explaining how something is used

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

candidate

-0.33

 tremp

-0.31

 deutsche

-0.29

 créations

-0.28

 ujarnya

-0.28

 garde

-0.28

 Yogyakarta

-0.28

 Tirol

-0.28

 AssemblyProduct

-0.28

 TextAppearance

-0.27

POSITIVE LOGITS



0.81

findpost

0.63

 beginnetje

0.62

majánló

0.59

 Chwiliwch

0.59

 estekak

0.58

 surla

0.57

ьажоргаш

0.57

 témoig

0.57

SBATCH

0.56

Activations Density 0.029%

expressions indicating personal experience or attempts at problem-solving

explaining how something is used

No Comments

No Known Activations