INDEX

Explanations

phrases indicating legal disclaimers or conditions

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

numbers and dates

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-65k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_65k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

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

Negative Logits

+#+#

-0.64

олові

-0.61

#+#

-0.59

Spoljašnje

-0.57

tagHelperRunner

-0.55

GeneratedCode

-0.54

 peix

-0.51

licação

-0.51

ьаж

-0.50

uebas

-0.49

POSITIVE LOGITS

Tikang

0.72

LookAnd

0.72

."]

0.61

〈

0.58

.].

0.58

bibfield

0.58

 ſta

0.57

Bum

0.57

 myſelf

0.57

"]:

0.56

Activations Density 0.004%

phrases indicating legal disclaimers or conditions

numbers and dates

No Known Activations