INDEX

Explanations

references to ingredients and instructions in recipes

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

and followed by action

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

 pleaſure

-0.63

 OFDb

-0.62

 ब्रेकडाउन

-0.60

 itſelf

-0.59

 ſtate

-0.58

 HasFactory

-0.57

featureID

-0.56

 Monfieur

-0.54

 Jefus

-0.54

 ſtand

-0.54

POSITIVE LOGITS

 hindurch

0.39

adpleegd

0.36

зулта

0.35

 ciudadana

0.34

 Савезне

0.34

়

0.33

Finish

0.33

 Comprometido

0.33

 Topf

0.32

 Nacionales

0.32

Activations Density 0.045%

references to ingredients and instructions in recipes

and followed by action

No Comments

No Known Activations