INDEX

Explanations

references to actions and roles in a process-oriented context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 7-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_7/width_16k/average_l0_69

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.7.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

 ftate

-0.99

 Majefty

-0.97

 ſtate

-0.97

 ſche

-0.97

 fevere

-0.95

 houſe

-0.95

 greateſt

-0.95

Eſ

-0.92

 bershka

-0.90

 neceff

-0.90

POSITIVE LOGITS

 nanti

0.60

 consideración

0.54

 nantinya

0.49

be

0.48

 todo

0.46

 Further

0.46

0.45

 future

0.45

Activations Density 0.398%

references to actions and roles in a process-oriented context

No Comments

No Known Activations