INDEX

Explanations

references to positional attributes and coordinates within a programming context

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

positionWS

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

Aholisi

-0.51

stridium

-0.39

exceptions

-0.37

 оригіналу

-0.36

AddRange

-0.35

 discutir

-0.34

طاق

-0.34

طار

-0.34

propylene

-0.34

bladet

-0.34

POSITIVE LOGITS

 position

0.91

 location

0.83

 Position

0.83

 POSITION

0.82

 위치

0.82

 positions

0.77

 位置

0.77

 ubicación

0.76

LLocation

0.76

位置

0.75

Activations Density 0.039%

references to positional attributes and coordinates within a programming context

positionWS

No Comments

No Known Activations