INDEX

Explanations

references to ethical concerns and exploitation

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

negativity or invalidity

np_acts-logits-general · gemini-2.0-flash

denying legitimate outcomes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 besonderer

-0.60

etcode

-0.57

 disponibilités

-0.57

 GenerationType

-0.56

šanai

-0.55

FormTagHelper

-0.54

 PyTuple

-0.52

行政区划

-0.52

 ulike

-0.52

السكان

-0.52

POSITIVE LOGITS

 innocent

1.42

 legitimate

1.41

 harmless

1.26

 perfectly

1.13

 innoc

1.13

 reputable

1.11

 valuable

1.10

innocent

1.09

 respectable

1.08

好不容易

1.06

Activations Density 0.697%

references to ethical concerns and exploitation

negativity or invalidity

denying legitimate outcomes

No Comments

No Known Activations