INDEX

Explanations

occurrences of the word "in."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 41-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_41/width_16k/average_l0_113

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.41.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

AndroidJUnit

-0.43

 intStringLen

-0.41

-0.38

ならない

-0.38

ftance

-0.38

 helplessly

-0.38



-0.36

 ActivatedRoute

-0.35

practice

-0.34

protoc

-0.33

POSITIVE LOGITS

 keeping

1.01

 line

0.90

 association

0.88

 conjunction

0.88

 alignment

0.85

 correlation

0.84

 cooperation

0.84

 collaboration

0.83

keeping

0.83

 conformity

0.82

Activations Density 0.579%

occurrences of the word "in."

No Comments

No Known Activations