INDEX

Explanations

elements related to structured data formats or typesetting commands

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-9b-pt-res/layer_20/width_16k/average_l0_68

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

posedge

-0.72

parsedMessage

-0.66

NUMX

-0.65

 endregion

-0.61

 Nieto

-0.59

BASEPATH

-0.57

SpringBootTest

-0.55

WSGI

-0.55

 INVENTION

-0.54

DeltaTime

-0.54

POSITIVE LOGITS

tabular

1.49

<table>

0.78

<tr>

0.65

mtable

0.59

 Inscrivez

0.56

hline

0.56

texttt

0.55

<thead>

0.53

}{|

0.52

FromArgb

0.50

Activations Density 0.008%

elements related to structured data formats or typesetting commands

No Comments

No Known Activations