INDEX

Explanations

instances of the word "describe" and variations thereof

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-9b-pt-res/layer_2/width_16k/average_l0_67

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.2.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

DockStyle

-0.58

 explanation

-0.57

 initComponents

-0.56

 intptr

-0.54

 EconPapers

-0.53

mannian

-0.45

før

-0.44

Compressor

-0.43

:✨

-0.42

berfläche

-0.42

POSITIVE LOGITS

describe

0.82

 describe

0.73

Describe

0.71

 Describe

0.69

 describes

0.61

 controversy

0.60

Descri

0.60

 Controversy

0.55

 Describes

0.54

Controversy

0.54

Activations Density 0.109%

instances of the word "describe" and variations thereof

No Comments

No Known Activations