INDEX

Explanations

phrases indicating conditional or introductory situations

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 37-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_37/width_16k/average_l0_124

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.37.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

RegressionTest

-0.65

afficheront

-0.59

TypedDataSet

-0.45

 utafitiHapana

-0.45

Дереккөздер

-0.44

 mostrarán

-0.44

下载附件

-0.43

にして

-0.43

neder

-0.43

 Inven

-0.42

POSITIVE LOGITS

 instance

2.05

 example

1.66

instance

1.58

 Instance

1.34

Instance

1.16

 instancia

1.13

 INSTANCE

1.04

 ejemplo

1.02

 eksempel

1.00

 exemplo

1.00

Activations Density 0.203%

phrases indicating conditional or introductory situations

No Known Activations