INDEX

Explanations

references to remaining quantities or counts in contexts involving categories or groups

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

remaining, rest, remainder

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

 finger

-0.40

 solas

-0.34

 carbonyl

-0.33

 calibrate

-0.32

MAE

-0.32

 identify

-0.32

 unify

-0.31

 validated

-0.31

 scoop

-0.31

 combat

-0.31

POSITIVE LOGITS

 iprot

0.69

RenderAtEndOf

0.69

SharedDtor

0.61

 autorytatywna

0.60

AnchorStyles

0.57

 übri

0.57

 antaranya

0.57

 restantes

0.56

runApp

0.55

enterOuterAlt

0.55

Activations Density 0.033%

references to remaining quantities or counts in contexts involving categories or groups

remaining, rest, remainder

No Comments

No Known Activations