INDEX

Explanations

numerical values and their formatting in the text

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-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

 betweenstory

-0.96

niſſe

-0.93

 InputDecoration

-0.93

ſſung

-0.93

iſchen

-0.91

 deſſen

-0.90

 témoig

-0.88

 vooz

-0.88

ðsíða

-0.87

 ſeines

-0.87

POSITIVE LOGITS

0.64

StandardCharsets

0.57

 charset

0.43

0.38

 civilización

0.33

0.32

UTF

0.32

0.31

 encoding

0.31

Activations Density 0.001%

numerical values and their formatting in the text

No Comments

No Known Activations