INDEX

Explanations

terms related to various scientific and technical concepts, particularly in chemistry and immunology

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

layer, protective, plane

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-res/layer_24/width_131k/average_l0_97

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.24.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

faßt

-0.36

はじめに

-0.36

 chances

-0.34

 tragically

-0.29

 Felsen

-0.29

 Teufel

-0.28

yg

-0.28

न्दर्भ

-0.28

jalá

-0.28

 froh

-0.27

POSITIVE LOGITS

niſſe

0.65

ſicht

0.63

IVEREF

0.63

ſcher

0.60

majánló

0.56

StructEnd

0.56

iſchen

0.56

GEBURTSDATUM

0.56

exitRule

0.55

Diwedd

0.54

Activations Density 0.157%

terms related to various scientific and technical concepts, particularly in chemistry and immunology

layer, protective, plane

No Comments

No Known Activations