INDEX

Explanations

numerical references and their relationships in a structured context

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-mlp-65k

Configuration

google/gemma-scope-2b-pt-mlp/layer_18/width_65k/average_l0_88

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.18.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

 Jefus

-1.12

Efq

-1.10

 myſelf

-1.09

 itſelf

-1.05

 Majefty

-0.99

ſelf

-0.99

 poffible

-0.98

 Monfieur

-0.96

 ſever

-0.94

 ſche

-0.92

POSITIVE LOGITS

aj

0.49

mo

0.48

vo

0.46

0.45

si

0.44

<eos>

0.43

 bruk

0.43

la

0.42

Activations Density 0.104%

numerical references and their relationships in a structured context

No Comments

No Known Activations