INDEX

Explanations

references to criticism and disapproval

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

critique and praise

np_acts-logits-general · gemini-2.0-flash

criticism and praise

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-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

fromnode

-0.42

JAKARTA

-0.40

 AppBundle

-0.40

Principal

-0.40

 uVar

-0.37

 Element

-0.37

 IFTT

-0.37

 Riggs

-0.37

get

-0.36

POSITIVE LOGITS

 Criticism

0.81

 criticism

0.81

Criticism

0.78

 criticizing

0.77

 criticisms

0.77

 criticized

0.75

 praising

0.75

 praise

0.73

 praised

0.72

 Praise

0.71

Activations Density 0.191%

references to criticism and disapproval

critique and praise

criticism and praise

No Comments

No Known Activations