INDEX

Explanations

references to influential figures and their actions or policies

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 23-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_23/width_16k/average_l0_120

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.23.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

-0.52

 Houſe

-0.48

 purpoſe

-0.46

 themſelves

-0.46

 bershka

-0.44

ACHUSET

-0.43

 defaultstate

-0.43

曖昧さ回避

-0.42

 оригіналу

-0.42

 pleaſure

-0.42

POSITIVE LOGITS

 that

0.61

 most

0.50

mà

0.42

IBinder

0.40

Dud

0.40

Linq

0.39

haer

0.39

ulti

0.38

 הכי

0.38

lida

0.38

Activations Density 0.170%

references to influential figures and their actions or policies

No Comments

No Known Activations