INDEX

Explanations

numerical identifiers and codes related to legal or administrative contexts

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

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

 sumpay

-0.73

 MainAxisSize

-0.63

ConstraintMaker

-0.60

providedIn

-0.59

testify

-0.59

 Autorisations

-0.58

 betweenstory

-0.55

 ostavi

-0.55

neux

-0.53

évaluateur

-0.52

POSITIVE LOGITS

 avoient

0.45

 étoient

0.43

 tár

0.37

 guste

0.34

 argint

0.33

 ifølge

0.31

Kaynakça

0.30

 placas

0.30

Viited

0.29

 Patienten

0.29

Activations Density 0.001%

numerical identifiers and codes related to legal or administrative contexts

No Comments

No Known Activations