INDEX

Explanations

phrases related to problems and problem-solving

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

problem solving

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

 içeris

-0.54

 mandiri

-0.54

ᶰ

-0.54

 Utilizamos

-0.53

Embrace

-0.51

 Embrace

-0.51

今でも

-0.51

 maiden

-0.50

今年は

-0.49

 dourada

-0.49

POSITIVE LOGITS

 problem

1.20

 Problem

1.19

problem

1.16

 problems

1.16

Problem

1.15

 PROBLEM

1.06

 Problems

1.05

problems

1.03

 Probl

1.02

PROBLEM

1.02

Activations Density 0.029%

phrases related to problems and problem-solving

problem solving

No Comments

No Known Activations