INDEX

Explanations

descriptions of experimental procedures and outcomes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 11-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_11/width_16k/average_l0_80

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.11.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

 Ward

-0.45

 concernant

-0.42

 által

-0.42

kan

-0.41

<eos>

-0.41

ichthys

-0.41

 körül

-0.41

nes

-0.40

 acontecimientos

-0.40

POSITIVE LOGITS

 Monfieur

0.96

 betweenstory

0.93

 ſeveral

0.89

 Majefty

0.89

 pleaſure

0.88

 Houſe

0.85

 myſelf

0.84

 auffi

0.84

 purpoſe

0.83

Personendaten

0.81

Activations Density 0.451%

descriptions of experimental procedures and outcomes

No Comments

No Known Activations