INDEX

Explanations

numerical identifiers associated with academic papers or articles

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 13-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_13/width_65k/average_l0_75

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.13.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

 copertura

-0.45

วย

-0.44

 tử

-0.43

gest

-0.43

voja

-0.42

posições

-0.42

americanas

-0.42

 základ

-0.41

uza

-0.40

estan

-0.40

POSITIVE LOGITS

IsContent

1.10

rungsseite

1.09

1.05

 createSlice

1.02

ArgsConstructor

1.01

enumii

1.00

 consultato

1.00

BufferException

0.97

StoryboardSegue

0.97

 NSCoder

0.96

Activations Density 0.013%

numerical identifiers associated with academic papers or articles

No Comments

No Known Activations