INDEX

Explanations

elements related to emotional response and aesthetic appreciation

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

[word] was/is/are [positive adjective]

np_acts-logits-general · gemini-2.0-flash

is bad

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 typelib

-0.75

MethodManager

-0.61

клопе

-0.54

才会

-0.54

存于互联网档案馆

-0.52

JNIEnv

-0.52

migrationBuilder

-0.51

setupUi

-0.51

HasColumnName

-0.51

carus

-0.51

POSITIVE LOGITS

is

1.03

 sucks

0.89

 είναι

0.81

 stinks

0.79

はとても

0.72

真是

0.71

 suck

0.66

 thật

0.66

 sucked

0.63

 feels

0.62

Activations Density 0.591%

elements related to emotional response and aesthetic appreciation

[word] was/is/are [positive adjective]

is bad

No Comments

No Known Activations