INDEX

Explanations

phrases indicating causation or results in various contexts

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

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Personendaten

-0.70

تقاوى

-0.63

 myſelf

-0.61

 >=",

-0.61

 بيها

-0.60

 pleaſure

-0.60

 /\.(

-0.59

 purpoſe

-0.59

 يتيمه

-0.57

SBATCH

-0.56

POSITIVE LOGITS

 result

1.49

result

1.27

 resultado

1.15

 product

1.10

 résultat

1.09

 результат

1.09

 producto

1.08

 Result

1.08

 consecuencia

1.07

 résult

1.05

Activations Density 0.457%

phrases indicating causation or results in various contexts

No Comments

No Known Activations