INDEX

Explanations

phrases that express absence or negation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 10-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_10/width_16k/average_l0_77

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.10.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

-1.12

 Monfieur

-0.99

 purpoſe

-0.87

ſelf

-0.85

 houſe

-0.82

 ſhe

-0.82

 quoque

-0.81

 Jefus

-0.80

 cauſe

-0.78

 pleaſure

-0.78

POSITIVE LOGITS

 Without

1.75

 without

1.69

Without

1.66

without

1.65

Ohne

1.55

WITHOUT

1.54

 WITHOUT

1.51

 ohne

1.39

 senza

1.30

Без

1.23

Activations Density 0.090%

phrases that express absence or negation

No Comments

No Known Activations