INDEX

Explanations

marker tokens indicating the beginning of a new section or entry

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

code or technical keywords

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 16-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_16/width_65k/average_l0_128

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.16.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

 myſelf

-1.03

 pleaſure

-1.01

 uſed

-1.01

:✨

-1.00

 deſt

-0.96

 raiſ

-0.95

 cauſe

-0.93

 CreateTagHelper

-0.92

 abſ

-0.91

 purpoſe

-0.90

POSITIVE LOGITS

<bos>

0.56

 getIs

0.47

 kind

0.43

 StatefulWidget

0.43

0.41

sonder

0.41

τζ

0.40

叫做

0.38

con

0.38

 that

0.38

Activations Density 0.106%

marker tokens indicating the beginning of a new section or entry

code or technical keywords

No Comments

No Known Activations