INDEX

Explanations

terminology and expressions related to algorithmic complexity

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 artists

-0.38

 TRAN

-0.38

itemize

-0.37

 Groot

-0.36

 Tran

-0.36

 bead

-0.36

 artist

-0.36

 Stone

-0.36

TRAN

-0.36

geben

-0.35

POSITIVE LOGITS

RenderAtEndOf

0.65

featureID

0.65

#+#

0.60

tagHelperRunner

0.56

 autorytatywna

0.54

MessageTagHelper

0.54

GEBURTSDATUM

0.53

تقاوى

0.52

iblichen

0.52

hoeddwyd

0.51

Activations Density 0.863%

terminology and expressions related to algorithmic complexity

No Comments

No Known Activations