INDEX

Explanations

statistical and factual reporting related to events and individuals

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

Negative Logits

featureID

-0.96

#+#

-0.79

 queſta

-0.73

 &___

-0.71

 increí

-0.71

majánló

-0.71

 indígen

-0.70



-0.70

richTextPanel

-0.68

stdc

-0.66

POSITIVE LOGITS

0.34

!!!

0.32

<strong>

0.32

<em>

0.32

 plan

0.29

kom

0.29

<h6>

0.29

 vand

0.29

¡

0.28

Plan

0.28

Activations Density 0.025%

statistical and factual reporting related to events and individuals

No Known Activations