INDEX

Explanations

expressions related to ease or difficulty in tasks or actions

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

unobstructed effortless ease

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 31-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_31/width_131k/average_l0_109

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.31.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

 EconPapers

-0.77

rungsseite

-0.71

 kasarigan

-0.59

 ModelExpression

-0.59

 informée

-0.58

 Infórmanos

-0.57

Personendaten

-0.57

 queſta

-0.54

+#+#

-0.52

tvguidetime

-0.52

POSITIVE LOGITS

 Easy

0.53

Easy

0.50

 easy

0.49

 effortless

0.47

 EASY

0.46

 breezy

0.46

 glide

0.44

 ease

0.44

easy

0.43

 Лег

0.41

Activations Density 0.008%

expressions related to ease or difficulty in tasks or actions

unobstructed effortless ease

No Comments

No Known Activations