INDEX

Explanations

temporal markers related to events

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

following "after" or "before"

np_acts-logits-general · gemini-2.0-flash

sequence marker after

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.18.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

 Cæsar

-0.90

 Houſe

-0.89

Efq

-0.85

 Monfieur

-0.84

 Shakspeare

-0.83

 Chriftian

-0.83

ſelf

-0.82

 greateſt

-0.82

 houſe

-0.81

 himſelf

-0.79

POSITIVE LOGITS

the

1.33

 they

1.24

it

1.20

 being

1.14

he

1.06

0.95

we

0.94

its

0.89

 having

0.88

an

0.81

Activations Density 0.169%

temporal markers related to events

following "after" or "before"

sequence marker after

No Comments

No Known Activations