INDEX

Explanations

concepts related to quality and accountability

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

preceding "itself" or "themselves"

np_acts-logits-general · gemini-2.0-flash

self-referential concepts and actions

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

 esetén

-0.98

NameInMap

-0.71

 ujednoznacz

-0.58

 Normdatei

-0.56

gameserver

-0.55

égard

-0.55

gendes

-0.54

Enders

-0.53

harusnya

-0.52

withId

-0.52

POSITIVE LOGITS

 itself

1.22

itself

1.01

 portion

0.97

 Itself

0.93

 themselves

0.92

 suivantes

0.91

iest

0.87

iniest

0.85

themselves

0.83

 aspect

0.82

Activations Density 2.710%

concepts related to quality and accountability

preceding "itself" or "themselves"

self-referential concepts and actions

No Comments

No Known Activations