INDEX

Explanations

actions involving conflict or confrontation

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 27-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_27/width_16k/average_l0_118

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.27.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

 sostenibilidad

-0.36

outez

-0.35

 Leidenschaft

-0.34

 duele

-0.32

 kaybet

-0.32

 sabemos

-0.32

 valiente

-0.31

 beğ

-0.30

রা

-0.29

 uitvo

-0.29

POSITIVE LOGITS

튿

0.69

parsedMessage

0.68

SequentialGroup

0.63

RTHOOK

0.61

 nahilalakip

0.60

 ſind

0.59

OGND

0.59

ſehen

0.59

 Chwiliwch

0.58

Efq

0.58

Activations Density 0.094%

actions involving conflict or confrontation

No Known Activations