INDEX

Explanations

alternative options or choices in a context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 0-gemmascope-mlp-65k

Configuration

google/gemma-scope-2b-pt-mlp/layer_0/width_65k/average_l0_72

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.0.hook_mlp_out

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

avax

-0.59

 Rechnung

-0.56

uisine

-0.55

-0.53

enchymal

-0.53

-0.52

valu

-0.52

acebook

-0.51

curity

-0.51

ENESS

-0.50

POSITIVE LOGITS

or

1.50

 или

1.25

或

1.09

 hoặc

1.04

或

1.02

 یا

1.02

 atau

1.02

 eller

1.01

หรือ

1.00

 또는

0.99

Activations Density 0.045%

alternative options or choices in a context

No Comments

No Known Activations