INDEX

Explanations

phrases related to alternatives or substitutions

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

in lieu of

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

 aimez

-0.52

asmussen

-0.50

mû

-0.49

 Noord

-0.46

 Travail

-0.44

yakit

-0.44

 derselbe

-0.44

NewReader

-0.44

 Zeller

-0.44

His

-0.44

POSITIVE LOGITS

 lieu

0.88

 Lieu

0.71

RenderAtEndOf

0.66

lieu

0.65

 замі

0.54

instead

0.52

脚注の使い方

0.52

Instead

0.50

 вместо

0.50

AnchorTagHelper

0.50

Activations Density 0.001%

phrases related to alternatives or substitutions

in lieu of

No Comments

No Known Activations